{"id":1150350,"date":"2025-11-04T09:57:20","date_gmt":"2025-11-04T17:57:20","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-academic-program&#038;p=1150350"},"modified":"2026-04-09T06:17:46","modified_gmt":"2026-04-09T13:17:46","slug":"microsoft-research-fellowship","status":"publish","type":"msr-academic-program","link":"https:\/\/www.microsoft.com\/en-us\/research\/academic-program\/microsoft-research-fellowship\/","title":{"rendered":"Microsoft Research Fellowship"},"content":{"rendered":"\n\n<p>Collaborating across disciplines to explore, innovate, and shape the future.<\/p>\n\n\n\n\n\n\n<h2 class=\"wp-block-heading\" id=\"about-the-program\">About the program<\/h2>\n\n\n\n<p>The Microsoft Research Fellowship creates opportunities for academic scholars (faculty, PhD students, and postdocs) to collaborate with Microsoft Research on open research challenges that advance scientific understanding, drive innovation, and deliver societal benefit.<\/p>\n\n\n\n<p>In an era of rapid innovation, this fellowship brings together academia and industrial researchers across disciplines to shape the future through open research collaborations. Academic collaboration is foundational to advancing scientific discovery, and today, it\u2019s more critical than ever. Academic institutions contribute deep domain expertise, novel perspectives, and a culture of exploration. Microsoft Research\u2019s global labs offer deep technical expertise, cutting-edge academic inquiry, and real-world application. By working together, we believe we can accelerate innovation, promote responsible development, and address complex challenges that can best be solved through interdisciplinary exploration.<\/p>\n\n\n\n\n\n<p>Microsoft is seeking scholars in academia who are pursuing research and instruction in academic venues across a range of disciplines.<\/p>\n\n\n\n<p>Eligibility guidelines vary by research challenge. Some challenges accept proposals from students, others from faculty, and some from both. Please refer to the \u201cResearch Challenges\u201d tab and the eligibility details below to understand the criteria for each challenge.<\/p>\n\n\n\n<p>This program accepts proposals from individuals based in the following regions: Africa, Australia & New Zealand, Canada, Europe, India, Hong Kong, Japan, South Korea, Singapore, Taiwan, Argentina, Brazil, Chile, Colombia, Mexico, Puerto Rico, and the United States.<\/p>\n\n\n\n<p>Microsoft actively seeks to foster greater levels of diversity in our workforce and in our pipeline of future researchers and collaborators. We are always looking for the best and brightest talent and celebrating individuality. We invite candidates to come as they are and do what they love.<\/p>\n\n\n\n<p><em>Fellows selected to be a part of the Microsoft Research Fellowship program are subject to disciplinary proceedings for inappropriate behavior, including but not limited to discrimination, harassment (including sexual harassment), or plagiarism will forfeit their funding. If accepted, a standard background screen will be conducted. Though Microsoft is not hiring you for employment, a successful screen is required as a stipulation to transfer funds under this fellowship.<\/em><\/p>\n\n\n\n\n\n<ul class=\"wp-block-list\">\n<li>Students may submit proposals on their own behalf. <em>See &#8220;How to Apply&#8221; for more details<\/em><\/li>\n\n\n\n<li>Students must be actively pursuing a PhD at a degree-granting, accredited institution<\/li>\n\n\n\n<li>Students must be enrolled at a degree-granting university in one of the following regions: Africa; Australia & New Zealand; Canada; Europe; India; Hong Kong; Japan; Korea; Singapore; Taiwan; Argentina; Brazil; Chile; Colombia; Mexico; Puerto Rico; or the United States.<\/li>\n<\/ul>\n\n\n\n\n\n<ul class=\"wp-block-list\">\n<li>Professors, Assistant Professors, or Associate Professors from a degree-granting, accredited institution may submit a proposal on their behalf. <em>See \u201cHow to Apply\u201d for more details<\/em><\/li>\n\n\n\n<li>Proposals submitted by eligible faculty should also identify a student (actively pursuing a PhD or relevant terminal degree) to be involved in the collaboration. If selected, both the faculty member and student will be named as Fellows. <\/li>\n\n\n\n<li>Faculty must have received their terminal degree (e.g., PhD, DSc). A terminal degree is defined as the highest-level college degree that can be achieved within your academic discipline or professional field.<\/li>\n\n\n\n<li>All faculty members and students listed in the proposal must be employed or enrolled at a degree-granting university located in one of the following regions: Africa; Australia & New Zealand; Canada; Europe; India; Hong Kong; Japan; Korea; Singapore; Taiwan; Argentina; Brazil; Chile; Colombi; Mexico; Puerto Rico; or the United States.<\/li>\n<\/ul>\n\n\n\n\n\n<ul class=\"wp-block-list\">\n<li>Postdocs may submit a proposal on their behalf. See \u201cHow to Apply\u201d for more details<\/li>\n\n\n\n<li>Proposals submitted by eligible postdocs should attest to support from a faculty advisor or department chair supporting postdoc&#8217;s participation in this collaboration with Microsoft through the Fellowship program.<\/li>\n\n\n\n<li>Postdocs must be employed or enrolled at a degree-granting university located in one of the following regions: Africa; Australia & New Zealand; Canada; Europe; India; Hong Kong; Japan; Korea; Singapore; Taiwan; Argentina; Brazil; Chile; Colombi; Mexico; Puerto Rico; or the United States.<\/li>\n<\/ul>\n\n\n\n\n\n<p>The Microsoft Research Fellowship is designed to create meaningful opportunities for collaboration between academic researchers and Microsoft Research. These opportunities will vary by research challenge and may include working sessions, regular check-ins, asynchronous collaboration, workshops, and other formats tailored to the nature of the research.<\/p>\n\n\n\n<p>All funds are distributed as unrestricted gifts as a one-time payment made directly to the institution\/university and disbursed according to the institution\/ university policies. Direct payments to individuals are not made by the program.<\/p>\n\n\n\n<p>Funding amounts vary by the region a fellows\u2019 organization\/institution of employment is located in.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Africa \u2013 $17,000 (USD)<\/li>\n\n\n\n<li>Australia & New Zealand \u2013 $17,000 (USD)<\/li>\n\n\n\n<li>Argentina, Brazil, Chile, Colombia, Mexico, Puerto Rico \u2013 $17,000 (USD)<\/li>\n\n\n\n<li>Canada \u2013 $47,000 (USD)<\/li>\n\n\n\n<li>Europe \u2013 $27,000 (USD)<\/li>\n\n\n\n<li>India \u2013 $17,000 (USD)<\/li>\n\n\n\n<li>Hong Kong, Japan, Korea, Singapore, and Taiwan \u2013 $17,000 (USD)<\/li>\n\n\n\n<li>United States \u2013 $47,000 (USD)<\/li>\n<\/ul>\n\n\n\n<p>These stipend amounts include travel funding to support in-person collaboration. Fellows may use this funding to visit Microsoft Research principal investigators and collaborators at a Microsoft Research lab or office. <\/p>\n\n\n\n<p>All payments will be distributed in March\/April 2026. If selected, Microsoft will work with respective departmental finance liaisons to transfer funds.<\/p>\n\n\n\n<p><em>*Stipend rates are reviewed annually and benchmarked against comparable fellowship programs in each geography to ensure competitive alignment.&nbsp;<\/em><\/p>\n\n\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n\n\n<h2 class=\"wp-block-heading\" id=\"research-challenges\">Research challenges<\/h2>\n\n\n\n<p>Research objectives and eligibility requirements vary by challenge. Some challenges accept proposals from PhD students, others from faculty, and some from both. Review the descriptions below to understand each challenge\u2019s research goals and specific eligibility criteria.<\/p>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<h4 class=\"wp-block-heading\" id=\"ai-for-global-and-societal-impact\">AI for global and societal impact<\/h4>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/cecilym\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Cecily Morrison<\/a> (Microsoft Research Cambridge, UK), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kahofman\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Katja Hofmann<\/a> (Microsoft Research Cambridge, UK)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/baym\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Nancy Baym<\/a> (Microsoft Research New England)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge explores the model capabilities and human scaffolds needed to enable creative professionals across geographies to meaningfully use and adopt interactive generative AI (interactive GenAI), including worlds models and multimodal models.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Interactive GenAI models generate novel virtual experiences, enabling users to move around and interact with spaces and artefacts that are generated in real time. It is now possible for media creatives to shape these experiences through the provision and curation of datasets for training, fine-tuning or prompting such models. Yet, we do not know what kinds of model capabilities and human scaffolds are needed to make novel interactive AI technologies useful for people in the creative media industry in ways that are equitable from the beginning. This project will explore this question through the concrete design brief of creating \u2018day-in-the-life\u2019 interactive GenAI experiences alongside disability advocacy organizations and the technical team developing interactive GenAI technologies. The resulting experiences will be released for public use, illustrating how people can shape AI.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We are looking for a creative artist who has technical skills to work with early stage pre-consumer technologies that will be a compelling demonstration of the possible for the general public. The candidate should have experience working with marginalized communities as well as enough technical expertise to interact fluidly with the technical team. Experience training and\/or fine-tuning generative AI models is ideal.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD students, faculty, and non-academia based proposals<\/p>\n\n\n\n<p><em>(Note: This research challenge is uniquely open to proposals from non-academic candidates. To submit, select \u201cFaculty\u201d in the portal in response to the profession field. Enter N\/A where a student name is requested. We\u2019ll recategorize your proposal on our end.) <\/em><\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/eldillon\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Eleanor Dillon<\/a> (Microsoft Research New England), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/liraymond\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Lindsey Raymond<\/a> (Microsoft Research New England)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: Eric Horvitz (Office of the Chief Scientific Officer)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>Combine firm-level data of GitHub Copilot adoption with detailed outside data on firms\u2019 employment, hiring, and performance to understand the impact of generative AI adoption on firm composition and employee outcomes, with a particular focus on off-shoring SDE work and differential impacts across countries.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>India produces about 1.5 million engineers every year, but survey and anecdotal evidence shows that many of them lack employable skills, including writing error-free code. For such workers, AI can potentially provide a powerful productivity increase, but also could substitute for some of their skills, altering the task composition of their work and consequently impacting wages. These impacts would likely vary by worker tenure as well as firm organizational structure. Furthermore, AI could help close the productivity gap between developed and developing country firms in certain sectors, but capital or managerial constraints may impede adoption of AI technologies. This project would use Github and other data to conduct exploratory research into potential interventions that would quantify how to optimize the introduction of AI tools into firms in India. We believe findings from this project could provide useful insights into firm decisions in developing countries broadly, given similar questions regarding employability of workers engaged in such occupations.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>This challenge is looking for a mid-career faculty member with expertise in the economics of organizations, technology adoption, and personnel management to collaborate on this research.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>Faculty proposals <\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jaoneil\/\">Jacki O\u2019Neill<\/a> (Microsoft Research Africa), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taganu\/\">Tanuja Ganu<\/a> (Microsoft Research India), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xingx\/\">Xing Xie<\/a> (Microsoft Research Asia)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ogbemie\/\">Ogbemi Ekwejunor-Etchie<\/a> (Microsoft Research Accelerator)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>The Global South AI Grand Challenge invites researchers to shape the future of AI from the Global South outward by defining new benchmarks, modalities, and solutions that drive progress on a global scale.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Artificial intelligence is reshaping how humanity learns, creates, and connects, yet its foundations remain incomplete, built on limited data, languages, and cultural perspectives. The Global South AI Grand Challenge invites researchers to expand the frontiers of AI by building from the majority of humanity outward.<\/p>\n\n\n\n<p>In partnership with Microsoft Research Africa, Microsoft Research India, Microsoft Research Asia and the Microsoft Research Accelerator, this challenge invites bold exploration in foundation and generative AI through novel datasets, architectures, training approaches, or model optimization for long-tailed data, all rooted in the cultural and local contexts of the Global South. Proposals may explore multilingual and multimodal intelligence, community-scale datasets, or novel evaluation frameworks that make AI more inclusive, adaptable, and trustworthy.<\/p>\n\n\n\n<p>Selected teams will collaborate with Microsoft researchers and engineers to test, refine, and scale their work on global platforms, advancing the shared goal of building AI systems that reflect and serve all of humanity.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We seek faculty with a strong track record of innovation in foundation models and generative AI, particularly those advancing video generation or model optimization for long tailed or low resource data. Preference will be given to researchers with a solid AI or ML background and demonstrated experience in cross-disciplinary collaboration that bridges technical and societal impact.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>Faculty proposals<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/asellen\/\">Abi Sellen<\/a> (Microsoft Research Cambridge UK), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/horvitz\/\">Eric Horvitz<\/a> (Office of the Chief Scientific Officer)<\/p>\n\n\n\n<p><strong>Additional Microsoft<\/strong> <strong>collaborators<\/strong>: Andrew Jenks (Microsoft), Jessica Young (Microsoft), and Sam Vaughn (Microsoft) &nbsp;&nbsp;<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge explores how we\u202fcan\u202fuse human-centric design to put the latest provenance tools (such as fingerprinting, watermarking, and cryptographic provenance technology) into the hands of users to allow them to understand and explore the source, history, and veracity of the online content they are interacting with.\u202f\u202f<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Tackling\u202fdisinformation and\u202fmisinformation is a growing and critical challenge. The rise of the generative capabilities of AI technologies to create and manipulate content threatens to usher in a \u201cpost-epistemic world,\u201d where fiction cannot be distinguished from reality. Microsoft has been a leader in the development of technical provenance tools. However, neither we nor others have invested enough in improving end-user experiences and understandings\u2014 or in assessing the effectiveness of designs for deploying these technical solutions. We seek a fellow to coordinate and focus cross-company efforts, while building connections to the broader ecosystem of organizations, on iterative human-centric efforts to build and evaluate methods for communicating provenance of content, with a specific focus on their effectiveness in improving end-user understanding, insight, and interpretation.<\/p>\n\n\n\n<p>The weakest link in helping end users understanding whether the content they are viewing was captured by cameras and microphones, produced through human effort, or generated wholly or partially by AI technologies is people\u2019s grasp of AI methods and their capabilities to create content and manipulate content, the intentions of content creators in different contexts, and the technical methods for marking content and encoding metadata.<\/p>\n\n\n\n<p>Thus, research on directions with managing disinformation and misinformation must be explicitly human-centered. A sociotechnical approach to the problem is critical.<\/p>\n\n\n\n<p>Such an approach starts with the assumption\u202fthat users need access to provenance information about digital content in order to\u202fmake sound judgements about what they are engaging with online.\u202f Rather than attaching simple labels to content (whether it be images, video, audio, or documents), the decision about whether something has been altered in ways that distort, deceive, or undermine its use, is personal, complex and highly contextual. We need social science research to understand the range of circumstances in which provenance information is important, the ways in which it is currently perceived or misperceived, and the factors that come into play when making these decisions. We also need design research to experiment with different ways that provenance information might be surfaced, and this effort will require iterative efforts with designing and testing different approaches.<\/p>\n\n\n\n<p>The aspiration is to design end user tools that would create new conventions across media to empower users to query and understand the origin of content and any alterations that have occurred, whether human or machine generated.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>This challenge is open to proposals from faculty, postdocs, and PhD students. We\u2019re looking for individuals with interdisciplinary experience in one or more of human-computer interaction, design, media and communications studies, security, and policy.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD student, postdoc, and faculty proposals<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n<\/div>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<h3 class=\"wp-block-heading h4\" id=\"dynamic-ui-for-collaborative-work\">AI fundamentals: scalable reasoning, model adaption and evaluation<\/h3>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wallach\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Hanna Wallach<\/a> (Microsoft Research NYC), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/mdudik\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Miro Dudik<\/a> (Microsoft Research NYC) <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xiaoyuanyi\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Xiaoyuan Yi <\/a>(Microsoft Research Asia)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xingx\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Xing Xie<\/a> (Microsoft Research Asia), Sociotechnical Alignment Center (Microsoft Research NYC), Societal AI group (Microsoft Research Asia)<\/p>\n\n\n\n<p><strong>Summary<\/strong>: This project aims to further bridge psychometrics and AI to develop a new science of Generative AI evaluation that moves beyond benchmarks toward interpretable, generalizable measures of model behavior.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Generative AI (GenAI) evaluation today relies heavily on benchmarks and leaderboards. Yet despite the proliferation of benchmarks purporting to capture diverse model capabilities and safety risks, it remains under what\u2014if anything\u2014benchmark scores individually or collectively tell us that generalizes beyond the specific tests on which they are reported. This research challenge seeks to catalyze a paradigm shift for GenAI evaluation by adapting and extending modern psychometric methods to meet the demands of this new domain. Through this challenge we aim to reconceptualize GenAI model capabilities, safety risks, and values as latent attributes of models that drive\u2014and hence are discoverable through\u2014observed AI behavior across diverse settings and use cases. This research challenge goes beyond simply applying existing psychometric frameworks and methods. Whereas data traditionally studied in psychometrics involves many test takers and a relatively small number of questions (a \u201clow-dimensional\u201d regime), in GenAI evaluation we have relatively few models which we subject to a battery of hundreds of thousands of questions (a \u201chigh-dimensional\u201d regime). The challenge aims to enhance existing psychometric methods through modern statistical machine learning to develop methods that are performant in the high-dimensional regime.\u202f<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>The ideal collaborators for this project would be a faculty member and PhD student with deep expertise in psychometrics, measurement, generative AI evaluation, and associated methods and theory from statistics and machine learning.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>Faculty and postdoc proposals<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/gauravsinha\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Gaurav Sinha<\/a> (Microsoft Research India), Kiran Shiragur (Microsoft Research India), and Shivam Garg (Microsoft Research AI Frontiers)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ariy\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Arun Iyer <\/a>(Microsoft Research India), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/someh\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Sonu Mehta<\/a> (Microsoft Research India)\u202f<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge aims to develop novel mechanisms for post-training retrieval models using high quality feedback from reward models (such as LLM based cross encoders) to optimize downstream retrieval and\/or generation (RAG) performance.\u202f<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Post-training has emerged as a powerful technique for steering language models toward maximizing desired rewards. This approach presents a significant opportunity for retrieval; a domain often constrained by sparse training data (i.e., lacking relevance signals for most query-document pairs). The availability of high-quality reward models, such as LLM-based cross-encoders or human feedback, makes this avenue particularly promising.\u202f<\/p>\n\n\n\n<p>We intend to develop and adapt post-training techniques specifically for retrieval models, targeting applications like Search, Advertising, and Retrieval-Augmented Generation (RAG).\u202f<\/p>\n\n\n\n<p>Some of our immediate research questions include:\u202f<\/p>\n\n\n\n<p>(1) Data Selection: Which queries and corresponding documents should be selected for scoring by reward models during the post-training stage to maximize downstream retrieval and RAG performance?\u202f<\/p>\n\n\n\n<p>(2) Loss Formulation: What is the optimal post-training loss function given a specific retrieval architecture, reward feedback design (e.g., pointwise, pairwise, listwise), and application scenario?\u202f<\/p>\n\n\n\n<p>(3) Computational Efficiency: How can we efficiently execute multiple post-training iterations, especially when the reward models are large and computationally expensive?\u202f<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>Ideal collaborators for this project include PhD students (and faculty) working in machine learning with a strong focus on reinforcement learning or information retrieval with an exposure to both theoretical and empirical research in these areas. Prior hands-on experience in working with large language models specifically in developing RL based post training algorithms will be extremely valuable. \u202f\u202f\u202f<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> PhD students and faculty <\/p>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/amshar\/\">Amit Sharma<\/a> (Microsoft Research India), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nagarajn\/\">Nagarajan Natarajan<\/a> (Microsoft Research India), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/niprasa\/\">Niranjani Prasad<\/a> (Microsoft Research Cambridge UK), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/skarmalkar\/\">Sushrut Karmalkar<\/a> (Microsoft Research Cambridge UK)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: Alicia Curth (Microsoft), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/vineethn\/\">Vineeth Balasubramanian<\/a> (Microsoft Research India)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge aims to advance knowledge-guided inference\u2014where external verifiers dynamically steer model generation at test time\u2014to unlock scalable, reliable reasoning in structured and high-stakes domains.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Scaling test-time compute has emerged as a key paradigm for reasoning in structured domains like math and code. This challenge explores new scaling dimensions through knowledge-guided inference\u2014a new paradigm where external verifiers (e.g., neuro-symbolic) and structured knowledge (e.g., domain-specific validators) actively steer model tokens during generation, going beyond the standard post-hoc verification paradigm. By scaling test-time compute through dynamic guidance, we aim to unify statistical learning with symbolic verification, enabling branching and backtracking correction during inference. A key sub-challenge is designing verifiers that can operate on ambiguous natural language, extending beyond formal domains like mathematics and code. This capability is essential for deploying reasoning models in high-stakes domains such as law and healthcare, where interpretability and reliability are critical.<\/p>\n\n\n\n<p>The collaboration aims to propose new verifier-guided architectures, training strategies for building verifiers in ambiguous scenarios, and practical mechanisms for deciding when to invoke external tools versus rely on internal reasoning. Ultimately, this work aims to redefine how reasoning models are built and deployed\u2014making them more adaptable, efficient, and trustworthy.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>Ideal collaborators for this project include PhD students (or faculty) in machine learning, natural language processing, or symbolic AI, with a strong focus on reasoning, verification, or neuro-symbolic methods. Expertise in reinforcement learning, building efficient AI inference systems, or program synthesis would be especially valuable.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD student and faculty proposals <\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n<\/div>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<h3 class=\"wp-block-heading h4\" id=\"dynamic-ui-for-collaborative-work\">Biological and scientific modeling<\/h3>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lualex\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Alex Lu <\/a>(Microsoft Research New England), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kevyan\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Kevin Yang <\/a>(Microsoft Research New England), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lcrawford\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Lorin Crawford<\/a> (Microsoft Research New England)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: Sarah Alamdari (Microsoft), Carles Domingo-Enrich (Microsoft), and <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ashleyconard\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Ashley Conard<\/a> (Health Futures)<\/p>\n\n\n\n<p><strong>Summary<\/strong>: Regulatory elements in non-coding DNA are critical to human health and genetic variation: can we use generative AI to understand and design them? \u202f<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>While generative models for biological sequence modalities (proteins, DNA) have exploded, the design of regulatory DNA remains elusive, due to the complex way these sequences interact in systems and their still-poorly understood nature. Unlocking this understanding would greatly expand human health, synthetic biology, and other applications, as it would allow for fine-grained control and understanding of the cell contexts genes in which are expressed. We wish to investigate if generative models of non-coding regulatory DNA can help us understand and design these elements. Themes of interest for us would include but not be limited to controlling cell type and state dependent regulation, applications to comparative genomics and bioinformatics, and design of de novo elements with properties not seen before in nature.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>The ideal collaborator would be a current faculty member who would be interested in using generative models to design and understand regulatory elements. We are seeking collaborators with a background in or related to regulatory genomics. Access to and experience in wet lab experimentation is considered a major asset.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>Faculty proposals <\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n<\/div>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<h3 class=\"wp-block-heading h4\" id=\"dynamic-ui-for-collaborative-work\">Foundational systems & infrastructure for AI<\/h3>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kstrauss\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Karin Strauss<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kalytv\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Kate Lytvynets <\/a>(Microsoft Research Redmond)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/bnguy\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Bichlien Nguyen<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jakesmith\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Jake Smith<\/a> (Microsoft Research Redmond), Danrong Zhang (Microsoft Research Redmond), <\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge is focused on developing and\/or evaluating the use of AI to address one or more of the many challenges in electricity planning and deployment.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Clean and affordable electricity is of critical importance for datacenters, their suppliers and the world. Deploying such electricity infrastructure is challenging, with a variety of chokepoints, including transmission bottlenecks, long interconnection and permitting queues, materials and siting challenges, policy challenges, supply constraints, and talent shortages. AI has recently emerged as a powerful tool to help with productivity, so this challenge involves developing new AI or using it to navigate these barriers\u2014accelerating grid modeling, optimizing siting decisions, forecasting renewable availability, and enabling new forms of coordination across the datacenter supply chain and energy ecosystem.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We are looking for interdisciplinary researchers, spanning areas such as computer science, electrical\/civil\/environmental engineering, policy, economics, etc.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD student, postdoc, and faculty proposals<\/p>\n\n\n\n<p><em>(Note: If you are a postdoc submitting a proposal for this challenge, please have faculty member submit the proposal, and enter your name and email in the \u201cStudent Name\u201d and \u201cStudent Email\u201d fields).<\/em> <\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\">\n\n<\/div>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/hiballan\/\">Hitesh Ballani<\/a> (Microsoft Research Cambridge UK), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/madanm\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Madan Musuvathi <\/a>(Microsoft Research Redmond)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/aashakashah\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Aashaka Shah<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/abonde\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Anand Bonde<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ishai\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Ishai Menache <\/a>(Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kmellou\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Konstantina Mellou<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/mmolinaro\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Marco Molinaro<\/a> (Microsoft Azure), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nswamy\">Nikhil Swamy<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/rdathathri\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Roshan Dathathri <\/a>(Microsoft Research Redmond),<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/saikatc\/?msockid=02b405ef82956e0f3d9f10f683226f80\"> Saikat Chakraborty<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sfakhoury\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Sarah Fakhoury <\/a>(Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sbarke\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Shraddha Barke<\/a> (Microsoft Research Redmond), Sirui Li (Microsoft Research Redmond)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge seeks to build a scalable, open-source ecosystem for reinforcement learning post-training, through foundational systems and algorithms research advances and unlock powerful reasoning capabilities for program intelligence, decision intelligence, and other rigorous, next-generation applications.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Reinforcement Learning (RL) based approaches have emerged as a critical component for post-training LLMs to enhance their reasoning abilities. Our goal is to build an open-source ecosystem to improve the efficiency and scalability of RL post-training, AI inference, and training by several orders of magnitude. This will allow efficient scaling of the post-training process to bigger models and larger datasets in the push towards achieving strong reasoning capabilities in program intelligence, optimization and decision intelligence, and other rigorous next-generation applications. To that end, we aim to:<\/p>\n\n\n\n<p>1) Exploit workload characteristics to develop system and algorithmic innovations across the stack, such as inference, training, network communication, and memory usage to improve scale by orders of magnitude.<\/p>\n\n\n\n<p>2) Explore existing algorithms and develop new algorithmic techniques to post-train MSR-specific models for cutting-edge applications.<\/p>\n\n\n\n<p>3) Leverage our symbolic reasoning expertise and large repositories of first-party and third-party software to unlock new program intelligence reasoning capabilities and power next-generation software engineering agents.<\/p>\n\n\n\n<p>4) Integrate optimization and generative AI tools to translate natural language descriptions into mathematical optimization problems, increase interpretability, and democratize access to advanced analytics tools.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>Describe the discipline, expertise, and area of focus you are seeking in a collaborator. Please be sure to confirm if you\u2019d like to target PhD students or faculty to collaborate with on your challenge.<\/p>\n\n\n\n<p>(1) Expertise: ML Systems, Systems, High-Performance Computing, GPU Programming and Optimizations, Program Reasoning, Compilers, Operations Research<\/p>\n\n\n\n<p>(2) Strong programming and experimentation skills, especially with LLM frameworks, and comfortable working with and debugging large-scale distributed ML systems<\/p>\n\n\n\n<p>(3) Currently pursuing a PhD in Computer Science or related fields<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD students and postdocs <\/p>\n\n\n\n<p><em>Note: Postdocs are invited to submit a proposal under the student track for this challenge. You may bypass the letter of recommendation upload process as this it not required for postdoc proposals. For questions, reach out to <a href=\"mailto:msfellow@microsoft.com\">msfellow@microsoft.com<\/a>.<\/em><\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n<\/div>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<h4 class=\"wp-block-heading\" id=\"tackling-misinformation-and-deepfakes-through-the-design-and-implementation-of-provenance-tools-for-end-users\">Human-AI collaboration and interaction<\/h4>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: Shannon Monroe (Microsoft Research Accelerator), Matt Corwine (Microsoft Research Accelerator), <\/p>\n\n\n\n<p><strong>Additional Microsoft Collaborators:<\/strong> Richard Banks (Microsoft Research Cambridge UK), Sean Rintel (Microsoft Research Cambridge UK), Neeltje Berger (Microsoft Research Accelerator)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>How can AI models better support creativity and innovation?<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>While AI models can help with idea generation and innovation, research shows that they tend to be homogeneous in nature &#8211; pushing creatives towards a common set of ideas and concepts. Innovation often comes about through the juxtaposition of two concepts that don&#8217;t look like they belong together, but when compared suggest new ideas and directions. Generative AI systems struggle with this context since their purpose is to weight items that belong together, rather than bring together those that don&#8217;t. How might we approach the design of new AI models whose goal is to operate right at the boundary of concepts that are connected and those that aren&#8217;t, in order to foster more radical forms of innovation?<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We\u2019re seeking a collaborator with expertise in computational creativity, cognitive science, and\/or generative AI systems, especially as they relate to human innovation. Ideal candidates will have a strong interdisciplinary orientation, a critical perspective on mainstream model design, and hands-on experience building or training AI models to explore conceptual boundaries and foster originality. They should be comfortable publishing in AI, HCI, or creativity-related venues and eager to challenge assumptions about how machines can support radical forms of innovation.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD student and faculty proposals<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sianl\/\">Si\u00e2n Lindley<\/a> (Microsoft Research Cambridge UK), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nath\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Nathalie Riche<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/johnwilliams\/\">Jack Williams<\/a> (Microsoft Research Cambridge UK), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nicmarquardt\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Nic Marquardt<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/romathugo\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Hugo Romat <\/a>(Microsoft Research Redmond)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This fellowship explores new abstractions and interaction principles for AI-powered collaborative dynamic experiences that adapt and morph to users\u2019 tasks and context.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>We aim to unlock a future where collaborative workflows are unconstrained by traditional software boundaries, and interfaces adapt and morph to users\u2019 tasks and context. Realizing this vision will require new computing abstractions, principles, and core interaction patterns, which empower users to work with these capabilities rather than be overwhelmed by them.<\/p>\n\n\n\n<p>This fellowship will focus on how dynamic experiences intersect with collaborative work. Research into generative UI (or malleable interfaces) tends to focus on individual experiences. However, modern work is inherently collaborative, and dynamic experiences will need to support collaboration if they are to have relevance. This might include the development of new systems to support AI-powered workflows such as team ideation and vibe coding, and latent collaboration tasks such as shared content creation.<\/p>\n\n\n\n<p>Key challenges in this space include (i) how to support users in collaboratively molding dynamic workspaces tailored to shared goals, (ii) how to support personalized representations for team members, which enable them to successfully collaborate while interacting with a shared underlying piece of work, and (iii) how to represent, integrate, and orchestrate multi-agent and multi-human collaboration across different configurations (human-human, human-agent(s), agent-agents acting on behalf of different humans) and across heterogenous interfaces.<\/p>\n\n\n\n<p>Working with the fellow, we hope to explore new experiences and articulate core principles for collaborative dynamic UI, enabling the balance of dynamism with consistent and shared mental models.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We seek to devise new interaction principles and abstractions for dynamic interaction with AI and develop proof-of concepts and prototypes to assess user value.<\/p>\n\n\n\n<p>A fellow should share our passion for Human-Computer Interaction (HCI), possess a solid understanding of fundamental design principles, and be familiar with recent advances in Human-AI Interaction such as generative user interfaces and malleable interaction patterns. Strong AI prototyping skills (including web stack development and experience with generative AI models or pipelines) are essential for demonstrating novel techniques and interaction patterns. Experience in conducting user studies to inform or assess user experiences would also be valuable.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD student, postdoc, and faculty proposals<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jenneville\/\">Jennifer Neville<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/blog\/tag\/sid-suri\/\">Sid Suri<\/a> (Microsoft Research Redmond), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kori\/\">Kori Inkpen<\/a> (Microsoft Research Redmond)<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sianl\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Sian Lindley <\/a>(Micrsoft Research Cambridge UK)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge explores how AI can move beyond individual assistance in complex knowledge work to become a true collaborator, teammate, and cognitive partner \u2014 through research in adaptive optimization and metrics, coordinated decision-making under uncertainty, preference elicitation and alignment, and information flow for collective intelligence.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>While current AI systems are optimized for individuals performing simple, independent tasks, organizational effectiveness depends on collaboration, coordination, and shared understanding across groups. As such, this challenge focuses on moving AI beyond narrow task assistance to functioning as a longer-horizon collaborator, teammate, and cognitive partner. We frame this problem as a multi-agent, continual-learning setting characterized by shifting goals, partial observability, and non-stationary rewards, where success requires more than the sum of individual contributions. Relevant research includes developing adaptive optimization and metrics that balance divergent stakeholder objectives, designing mechanisms for coordinated decision-making under uncertainty, advancing methods for preference elicitation and alignment, and improving information flows that sustain collective intelligence. We aim to amplify both individual productivity and organizational effectiveness, while deepening connections between AI\/ML research and social-organizational theory. Tangible outcomes include new algorithms for adaptive optimization and coordination, benchmarks and evaluation protocols that capture group-level objectives, and prototypes of AI systems capable of enhancing collective reasoning, decision-making, and productivity in real-world knowledge-work settings.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>This research challenge invites applicants from both AI\/ML and the social sciences who are interested in advancing AI as a collaborator, teammate, and cognitive partner in complex knowledge work. Ideal collaborators may be AI\/ML researchers with expertise in areas such as adaptive optimization, multi-agent systems, or alignment, or computational social scientists (including those in HCI and cognitive science) who study coordination, communication, and collective intelligence. Regardless of background, we are especially seeking researchers who are eager to work across disciplines to connect technical advances with data-driven insights into human and organizational dynamics.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>Faculty and postdoc proposals<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jialia\/\">Jianxun Lian<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dongsli\/\">Dongsheng Li<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xingx\/\">Xing Xie<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/bainguo\/\">Baining Guo<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/besh\/\">Beibei Shi<\/a> (Microsoft Research Asia)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>Develop and evaluate socially intelligent AI agents that model human cognition, coordinate with other agents, and collaborate productively with people to solve complex real-world tasks in education, science, and organizational operations.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>We aim to build socially intelligent AI agents that can understand, predict, and respond to human cognition and behavior while coordinating with other agents. Despite recent advances, today\u2019s systems lag behind humans in theory of mind, communication, negotiation, and sustained collaboration, constraining their effectiveness in complex workflows. We invite participants to develop rigorous, interdisciplinary methods \u2014 grounded in psychology, cognitive science, HCI, and organizational behavior \u2014 for systematically investigating and enhancing social reasoning and multi-agent collaboration. Approaches should incorporate mechanisms for trust, safety, accountability, and aligned goal pursuit in human-AI teams. Target domains include classroom tutoring and group learning, collaborative scientific discovery, and organizational operations such as planning, decision-making, and project execution. Expected outcomes include open-source models and toolkits, curated datasets and benchmarks, empirical studies and research publications, and deployable prototypes integrated into real workflows. The results will also help frontier firms and organizations prepare for the next wave of effective, safe human-AI collaboration at scale.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We seek collaborators across various research domains (LLMs, multi-agent RL, social sciences, HCI, etc.) who specialize in social reasoning, multi-agent coordination, and human-AI teaming. Ideal partners can design rigorous experiments, run human-subject and field studies, and build deployable prototypes in education, science, or organizational settings. We welcome faculty, PhD students, and postdocs, with preference for interdisciplinary teams and access to user populations or enterprise workflows.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>Faculty, PhD student, and postdoc proposals <\/p>\n\n\n\n<p><em>(Note: If you are a PhD student or postdoc, please have the involved faculty member submit the proposal, and enter your name and email in the \u201cStudent Name\u201d and \u201cStudent Email\u201d fields).<\/em> <\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n<\/div>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<h4 class=\"wp-block-heading\" id=\"multimodal-embodied-intelligence\">Multimodal & Embodied Intelligence<\/h4>\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/vineethn\/\">Vineeth N Balasubramanian<\/a> (Microsoft Research India), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taganu\/\">Tanuja Ganu<\/a> (Micosoft Research India),<\/p>\n\n\n\n<p><strong>Additional Microsoft collaborators<\/strong>: <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/www.linkedin.com\/in\/mercypeter\/?originalSubdomain=in\">Mercy Ranjit <span class=\"sr-only\"> (opens in new tab)<\/span><\/a>(Microsoft Research India) <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/neeraka\/\">Neeraj Kayal<\/a> (Microsoft Research India), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ogbemie\/\">Ogbemi Ekwejunor-Etchie<\/a> (Microsoft Research Accelerator)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge explores foundational multimodal LLM architectures that move beyond tokenization, aligning structurally and functionally with modality-specific characteristics \u2014 drawing inspiration from human cognition to enable inclusive, efficient, and robust reasoning across modalities like vision, speech, and action.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Contemporary large language models have made remarkable progress in text understanding but remain limited in how they process and reason across other modalities such as images, video, and speech. Current multimodal approaches typically extend text-based tokenization pipelines to other modalities, which fails to capture their unique structural and relational properties. This challenge aims to explore foundational architectures for multimodal LLMs that move beyond tokenization toward representations that are structurally and functionally aligned with each modality\u2019s characteristics. Inspired by principles of human cognition, where distinct sensory regions integrate information in complementary ways, we seek to design modular and adaptive components that enable more natural cross-modal understanding and reasoning. The research will investigate how these architectures can interleave attention across modalities, improving both perception and inference capabilities in complex, real-world tasks. A particular focus will be on efficiency, achieving parameter- and sample-efficient learning while maintaining strong generalization across modalities. The outcomes will include prototype architectures, evaluation benchmarks, and multimodal reasoning pipelines applicable to domains such as robotics, copilots, and embodied AI systems. Beyond advancing foundational AI research, this research challenge has the potential to unlock inclusive technologies for the global majority, where much of the world\u2019s knowledge exists in non-textual forms such as video and spoken interaction.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We seek collaboration with faculty and PhD students in AI, machine learning, or computer vision with expertise in multimodal representation learning, efficient model architectures, and large-scale foundation model training. Ideal collaborators will have a strong research track record in areas such as vision-language models, transformer architectures, as well as building and evaluating multimodal systems. Experience in cognitive-inspired AI modeling would be desirable but not mandatory. We aim to jointly explore foundational advances in multimodal LLM architectures that go beyond tokenization, with potential for long-term academic-industry impact.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD student, postdoc, and faculty proposals<\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jiaoyan\/\">Jiaolong Yang<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lizo\/\">Li Zhao<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jiabia\/\">Jiang Bian<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/bainguo\/\">Baining Guo<\/a> (Microsoft Research Asia), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lisu\/\">Lily Sun<\/a> (Microsoft Research Accelerator), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jfgao\/?msockid=02b405ef82956e0f3d9f10f683226f80\">Jianfeng Gao<\/a> (Microsoft Research Redmond)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>Develop and advance foundation models that empower robots to perform a wide variety of tasks with flexibility, reliability, and adaptability across real-world environments.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Foundation models have shown transformative potential in language and vision domains, and their application in robotics is an exciting frontier for artificial intelligence. This challenge seeks research into designing, training, and evaluating general-purpose robotics foundation models that support perception, reasoning, and action in diverse scenarios. We welcome work on model architectures, scalable training techniques, and approaches that enhance generalization and robustness for robots operating in complex, dynamic environments. Possible directions include leveraging multi-modality inputs and internet-scale data, fast adaptation to varied tasks, safety, and recovery from failure. Successful collaborations might deliver new model designs, open-source foundation models, or robust evaluation benchmarks benefiting the robotics research community. The goal is to facilitate the development of robots that can quickly learn new skills, adapt to new situations, and execute tasks with reliability. Long-term impact spans improved workflows, safer and more capable robots for applications across industry, healthcare, and daily life.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We are seeking faculty collaborators (professors or principal investigators) with strong backgrounds in robotics, computer vision, multimodal AI, and reinforcement learning. Experience with foundation models such as LLM\/VLM\/VLA\/VideoGen and real-world robotics systems is highly desirable. We welcome proposals from faculty who are eager to bridge AI research and practical robotics applications in partnership with Microsoft Research.<\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>Faculty proposals <\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n\n\n\n\n\n<p><strong>Principal Investigator(s)<\/strong>: <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/www.linkedin.com\/in\/mikepmurray\/\">Michael Murray<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research Accelerator), <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" rel=\"noopener noreferrer\" target=\"_blank\" href=\"https:\/\/www.linkedin.com\/in\/thellebrekers\/\">Tess Hellebrekers<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research Accelerator), <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tanreuben\/\">Reuben Tan<\/a> (Microsoft Research Redmond)<\/p>\n\n\n\n<p><strong>Summary<\/strong>:<\/p>\n\n\n\n<p>This challenge explores how robots can leverage alternative data sources including human video, simulation, and synthetic augmentation to scale learning without relying solely on expensive teleoperated demonstrations.<\/p>\n\n\n\n<p><strong>Description<\/strong>: <\/p>\n\n\n\n<p>Current robot learning approaches heavily depend on teleoperated demonstrations, which are expensive to collect and difficult to scale across diverse tasks and environments. This challenge seeks innovative methods to enable robots to learn from alternative data sources that are more readily available or easier to generate at scale. Key data sources of interest include human video demonstrations, physics simulations, procedurally generated synthetic data, and augmented variations of limited real robot data. Successful approaches will need to address fundamental challenges in domain adaptation, including differences in embodiment between humans and robots, the sim-to-real gap, and distribution shift between training and deployment environments. We are particularly interested in methods that can effectively combine multiple data modalities, leverage pretrained vision-language models, and develop robust representations that transfer across domains. The ultimate goal is to achieve robot learning systems that match or exceed the performance of teleoperation-trained policies while requiring orders of magnitude less robot-specific data collection. Solutions should demonstrate generalization to novel objects, tasks, and environments beyond the training distribution. This research has the potential to dramatically accelerate the deployment of capable robots in real-world applications from manufacturing to home assistance.<\/p>\n\n\n\n<p><strong>Ideal collaborator<\/strong>:<\/p>\n\n\n\n<p>We seek researchers with expertise in robot learning, computer vision, domain adaptation, or simulation who are passionate about making robot learning more practical and scalable. Ideal collaborators will have experience with imitation learning, reinforcement learning, or transfer learning approaches, and familiarity with modern deep learning frameworks and robot simulation environments. We particularly value creative problem-solvers who can bridge the gap between different data modalities and have a track record of developing methods that work reliably in real-world robotic systems.<\/p>\n\n\n\n<p>This challenge is only open to proposals from PhD students. <\/p>\n\n\n\n<p><strong>Eligible candidates:<\/strong> <\/p>\n\n\n\n<p>PhD students <\/p>\n\n\n\n<div class=\"wp-block-buttons is-layout-flex wp-block-buttons-is-layout-flex\"><\/div>\n\n\n<\/div>\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<h2 class=\"wp-block-heading\" id=\"how-to-apply\">How to apply<\/h2>\n\n\n\n<p>Please carefully review and consider the below as it relates to the proposal process for the program.&nbsp;&nbsp;<\/p>\n\n\n\n<p><strong>Proposal criteria<\/strong><\/p>\n\n\n\n<p>All proposals should be submitted via the submission portal (below). Proposals will not be accepted by email.&nbsp;<\/p>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33.33%\">\n<h3 class=\"wp-block-heading\" id=\"key-dates\">Key dates:<\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Call for proposals: <\/strong>Tuesday, November 4, 2025<\/li>\n\n\n\n<li><strong>Proposal submission deadline: <\/strong>Tuesday, December 16, 2025, at 11:59 PM ET <em>(the submission process is now closed)<\/em>.<\/li>\n\n\n\n<li><strong>Proposal notifications sent<\/strong>: February 27, 2026<\/li>\n<\/ul>\n<\/div>\n<\/div>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"general-guidelines\">General guidelines<\/h4>\n\n\n\n<ul class=\"wp-block-list\">\n<li>All submissions should be submitted in English by the candidate (self-submission).&nbsp;Incomplete applications will not be considered.&nbsp;<\/li>\n\n\n\n<li>Please do not add additional requests for budget, as these will not be considered under the guidelines of this fellowship.&nbsp;<\/li>\n\n\n\n<li>Proposals submitted to Microsoft will not be returned. Microsoft cannot assume responsibility for the confidentiality of information in submitted applications. Therefore, proposals should not contain information that is confidential, restricted, or sensitive.&nbsp;<\/li>\n\n\n\n<li>Due to the volume of submissions, Microsoft Research cannot provide individual feedback on applications that do not receive fellowship awards.&nbsp;<\/li>\n<\/ul>\n\n\n\n\n\n<p><em>Review eligibility guidelines on the Overview tab before proceeding with your proposal. <\/em><\/p>\n\n\n\n<p><strong>Each submission will include<\/strong>:&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Student name <\/li>\n\n\n\n<li>University where the student is actively enrolled<\/li>\n\n\n\n<li>Department <\/li>\n\n\n\n<li>Region and country where student\u2019s university is located<\/li>\n\n\n\n<li>Research focus summary (100 words or less) <\/li>\n\n\n\n<li>\u201cStatement of Interest\u201d title (80 characters or less)&nbsp;<\/li>\n\n\n\n<li>Submission of a \u201cStatement of Interest\u201d document. See additional instructions and guidance in the \u201cStatement of Interest Requirements\u201d section below.&nbsp;<\/li>\n\n\n\n<li>Candidate CV\/resume to be uploaded in the submission portal.&nbsp;<\/li>\n\n\n\n<li>Letters of Recommendation Requirement\n<ul class=\"wp-block-list\">\n<li>All PhD students submitting a student-led proposal must include two letters of recommendation.<\/li>\n\n\n\n<li>Guidance for Letters of Recommendation\n<ul class=\"wp-block-list\">\n<li>Students will be asked to request two signed letters on institutional or professional letterhead.<\/li>\n\n\n\n<li>Each letter should be uploaded as a separate PDF file.<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li>Letter of Recommendation 1 of 2: This letter must be from your faculty advisor or research supervisor who can speak to your academic work and research qualifications. <\/li>\n\n\n\n<li>Letter of Recommendation 2 of 2. This second letter is at your discretion. It may come from another faculty member, mentor, or professional who knows your work well.<\/li>\n\n\n\n<li><strong>Update: Two Options to Submit Letters of Recommendation<\/strong>\n<ul class=\"wp-block-list\">\n<li>The submission process will allow for confidential letter of recommendation submissions. Students will have two choices:\n<ul class=\"wp-block-list\">\n<li>1. Upload the letter themselves via their proposal form <\/li>\n\n\n\n<li>2. Provide the recommender\u2019s name and email in the proposal submission form. Our program will then contact the recommender (after the Call for Proposals deadline closes) with instructions to submit the letter directly and confidentially. Recommenders will have until January 8th to submit their Letter of Recommendation. For questions on this process, contact at msfellow@microsoft.com.<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<p><strong>A strong submission will<\/strong>:&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Demonstrate a clear understanding of the research challenge and its aims for scientifical, technological, or societal impact.<\/li>\n\n\n\n<li>Describe relevant work\/accomplishments, demonstrate the alignment one\u2019s motivation to the research challenge, and outline the potential impact achieve through collaboration.<\/li>\n\n\n\n<li>Be clearly written, in alignment with program guidelines.<\/li>\n\n\n\n<li>Consider that readability and clarity are valued in the review process.<\/li>\n<\/ul>\n\n\n\n<p><strong>Statement of Interest guidelines<\/strong>: <\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Describe the motivation, or \u201cwhy\u201d, behind candidate\u2019s interest in research exploration in this research challenge. We are specifically looking for candidates to describe how their research goals and aspirations align to this challenge.&nbsp;<\/li>\n\n\n\n<li>Description of relevant qualifications, expertise and perspectives the candidate would bring to the research challenge and how this experience would impact the outcomes of this collaboration.&nbsp;<\/li>\n\n\n\n<li>Detail any relevant areas that you would like to collaborate on with the Microsoft researchers and applied scientists listed as involved in the research challenge and how collaboration could create impact in those areas.\u202f<\/li>\n\n\n\n<li>Where applicable, include details of any relevant, preliminary research or work that shows progress and investment in this space.<\/li>\n\n\n\n<li>Statement of Interest documents should be a minimum 10-point font with a maximum of 3 pages. You may include a fourth page with citations and references as part of your Statement of Interest. The total length of this document should not exceed 4 pages total, assuming the fourth page is dedicated to citations\/references only. Margins should be 1\u201d or wider to ensure readability. Option to submit as word document or PDF file.&nbsp;<\/li>\n<\/ul>\n\n\n\n\n\n<p><em>Review eligibility guidelines on the Overview tab before proceeding with your proposal. <\/em><\/p>\n\n\n\n<p><strong>Each submission will include<\/strong>:&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Postdoc name<\/li>\n\n\n\n<li>University or institution where the postdoc is actively employed<\/li>\n\n\n\n<li>Department <\/li>\n\n\n\n<li>Region and country where postdoc&#8217;s university or institution is located<\/li>\n\n\n\n<li>Research focus summary (100 words or less) <\/li>\n\n\n\n<li>\u201cStatement of Interest\u201d title (80 characters or less)&nbsp;<\/li>\n\n\n\n<li>Submission of a \u201cStatement of Interest\u201d document. See additional instructions and guidance in the \u201cStatement of Interest Requirements\u201d section below.&nbsp;<\/li>\n\n\n\n<li>Candidate CV\/resume to be uploaded in the submission portal.&nbsp;<\/li>\n\n\n\n<li>Candidate&#8217;s will be required to confirm they have support from a faculty advisor or department chair supporting postdoc&#8217;s participation in this collaboration with Microsoft through the Fellowship program.<\/li>\n<\/ul>\n\n\n\n<p><strong>A strong submission will<\/strong>:&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Demonstrate a clear understanding of the research challenge and its aims for scientifical, technological, or societal impact.<\/li>\n\n\n\n<li>Describe relevant work\/accomplishments, demonstrate the alignment one\u2019s motivation to the research challenge, and outline the potential impact achieve through collaboration.<\/li>\n\n\n\n<li>Be clearly written, in alignment with program guidelines.<\/li>\n\n\n\n<li>Consider that readability and clarity are valued in the review process.<\/li>\n<\/ul>\n\n\n\n<p><strong>Statement of Interest guidelines<\/strong>: <\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Describe the motivation, or \u201cwhy\u201d, behind candidate\u2019s interest in research exploration in this research challenge. We are specifically looking for candidates to describe how their research goals and aspirations align to this challenge.&nbsp;<\/li>\n\n\n\n<li>Description of relevant qualifications, expertise and perspectives the candidate would bring to the research challenge and how this experience would impact the outcomes of this collaboration.&nbsp;<\/li>\n\n\n\n<li>Detail any relevant areas that you would like to collaborate on with the Microsoft researchers and applied scientists listed as involved in the research challenge and how collaboration could create impact in those areas.\u202f<\/li>\n\n\n\n<li>Where applicable, include details of any relevant, preliminary research or work that shows progress and investment in this space.<\/li>\n\n\n\n<li>Statement of Interest documents should be a minimum 10-point font with a maximum of 3 pages. You may include a fourth page with citations and references as part of your Statement of Interest. The total length of this document should not exceed 4 pages total, assuming the fourth page is dedicated to citations\/references only. Margins should be 1\u201d or wider to ensure readability. Option to submit as word document or PDF file.&nbsp;<\/li>\n<\/ul>\n\n\n\n\n\n<p><em>Review eligibility guidelines on the Overview tab before proceeding with your proposal. <\/em><\/p>\n\n\n\n<p><strong>Each submission will include<\/strong>:&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Name of faculty member <\/li>\n\n\n\n<li>University\/institution of employment <\/li>\n\n\n\n<li>Department<\/li>\n\n\n\n<li>Professional title (Professor, Associate Professor, or Assistant Professor) <\/li>\n\n\n\n<li>Region and country of employment <\/li>\n\n\n\n<li>Identify a student to participate in collaboration (first, last name and email address will be required) <\/li>\n\n\n\n<li>Research focus summary (100 words or less) <\/li>\n\n\n\n<li>\u201cStatement of Interest\u201d title (80 characters or less)&nbsp;<\/li>\n\n\n\n<li>Submission of a \u201cStatement of Interest\u201d document. See additional instructions and guidance in the \u201cStatement of Interest Requirements\u201d section below.&nbsp;<\/li>\n\n\n\n<li>Candidate CV\/resume to be uploaded in the submission portal.&nbsp;<\/li>\n<\/ul>\n\n\n\n<p><strong>A strong submission will<\/strong>:&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Demonstrate a clear understanding of the research challenge and its aims for scientifical, technological, or societal impact.<\/li>\n\n\n\n<li>Describe relevant work\/accomplishments, demonstrate the alignment one\u2019s motivation to the research challenge, and outline the potential impact achieve through collaboration.<\/li>\n\n\n\n<li>Be clearly written, in alignment with program guidelines.&nbsp;<\/li>\n\n\n\n<li>Consider that readability and clarity are valued in the review process.&nbsp;<\/li>\n<\/ul>\n\n\n\n<p><strong>Statement of Interest guidelines<\/strong>: <\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Describe the motivation, or \u201cwhy\u201d, behind candidate\u2019s interest in research exploration in this research challenge. We are specifically looking for candidates to describe how their research goals and aspirations align to this challenge.&nbsp;<\/li>\n\n\n\n<li>Description of relevant qualifications, expertise and perspectives the candidate would bring to the research challenge and how this experience would impact the outcomes of this collaboration.&nbsp;<\/li>\n\n\n\n<li>Detail any relevant areas that you would like to collaborate on with the Microsoft researchers and applied scientists listed as involved in the research challenge and how collaboration could create impact in those areas.\u202f<\/li>\n\n\n\n<li>Where applicable, include details of any relevant, preliminary research or work that shows progress and investment in this space.<\/li>\n\n\n\n<li>Description of the student identified for the collaboration, highlighting their academic background, specific expertise, and how their work relates to the research challenge the proposal aligns with. <\/li>\n\n\n\n<li>Statement of Interest documents should be a minimum 10-point font with a maximum of 3 pages. You may include a fourth page with citations and references as part of your Statement of Interest. The total length of this document should not exceed 4 pages total, assuming the fourth page is dedicated to citations\/references only. Margins should be 1\u201d or wider to ensure readability. Option to submit as word document or PDF file.&nbsp;<\/li>\n<\/ul>\n\n\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n\n\n<h2 class=\"wp-block-heading\" id=\"frequently-asked-questions\">Frequently asked questions<\/h2>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"proposals\">Proposals<\/h3>\n\n\n\n\n\n<p>Please visit How to Apply tab detailed information on what should be included in your proposal.<\/p>\n\n\n\n\n\n<p>A link to the submission portal is available at the top of the How to Apply\u202ftab. Proposals submitted via email will not be accepted.<\/p>\n\n\n\n\n\n<p>Your Statement of Interest is a document required as part of the proposal process. The document should use a minimum 10-point font and be no more than three pages in length. You may include a fourth page for citations and references. Use margins of at least 1 inch for readability. You may submit your proposal as either a Word document or PDF.<\/p>\n\n\n\n\n\n<p>Each candidate may submit up to three proposals for the fellowship program. You may submit only one proposal per challenge. Submitting multiple proposals to the same challenge will result in the second proposal being rejected.<br>The submission portal allows one proposal per email address. If you choose to submit a proposal for a different challenge, you will need to use a different email address for that submission. If accepted, you will have the opportunity to update your email on file to reflect your preferred email. Each proposal will be evaluated separately by the Microsoft Research PIs for the respective challenge. You may only be accepted under one research challenge. <\/p>\n\n\n\n\n\n<p>Yes, as long as your project aligns with the research directions outlined in the Research Challenge section. Proposals should include a Statement of Interest that addresses the criteria listed in the How to Apply section.<\/p>\n\n\n\n\n\n<p>The fellowship will encompass an open research collaboration between fellows and Microsoft, beginning in March 2026 and concluding in June 2027. The nature and structure of collaboration will vary by research challenge. This will all be coordinated once fellows are selected and during the initial discussions with Microsoft Principal Investigators and collaborators.<\/p>\n\n\n\n\n\n<p>No, there is no limit. All candidates who are eligible are welcome to submit a proposal.<\/p>\n\n\n\n\n\n<p>The Fellowship program does not support joint proposals at this time.<\/p>\n\n\n\n\n\n<p>For challenges open to <em>both <\/em>student-led and faculty-led proposals.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>A PhD student may submit a proposal independently for this challenge. All student-led proposals require faculty approval (meaning the faculty member supports the student investing time in this collaboration).<\/li>\n\n\n\n<li>This challenge also accepts faculty-led proposals. You&#8217;ll see that this proposal type requires the faculty member to name a student collaborator as a part of the proposal. In this scenario, both the faculty member and the PhD student (if selected) would be named Fellows and active participants in the collaboration.<\/li>\n<\/ul>\n\n\n\n<p>While you may submit one of each, we recommend the faculty member and student consider how they wish to participate and choose the path that best aligns with their goals and interests.<\/p>\n\n\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"review-selection\">Review & selection<\/h3>\n\n\n\n\n\n<p>Proposals will be reviewed by the Microsoft PIs and collaborators listed on the Research Challenges tab. For evaluation criteria, please refer to the How to Apply section. In general, reviewers will consider:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Research accomplishments<\/li>\n\n\n\n<li>Alignment with the selected research challenge<\/li>\n\n\n\n<li>Potential impact of the proposed collaboration<\/li>\n<\/ul>\n\n\n\n\n\n<p>All applicants will be notified of their status by February 2026. Due to the volume of submissions, we are unable to provide individual feedback.<\/p>\n\n\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"provisions-of-the-award-1\">Provisions of the award<\/h3>\n\n\n\n\n\n<p>Selected candidates will be named Fellows and receive a monetary award in the form of an unrestricted gift to their institution, intended to serve as a stipend. Fellows will also collaborate with Microsoft researchers on the selected challenge. This award does not affect eligibility for future internships or employment at Microsoft, nor does it increase the likelihood of receiving them. Fellows may also accept funding from other organizations.<\/p>\n\n\n\n\n\n<p>Fellowship awards will be distributed in March\/April 2026 via electronic payment to the recipient\u2019s institution. Upon acceptance, fellows will be asked to connect Microsoft with their institution\u2019s finance department to coordinate payment.<\/p>\n\n\n\n\n\n<p>The tax implications for the fellowship are based on the policy at your university and applicable tax laws in your region.<\/p>\n\n\n\n\n\n<p>The funds are given as an unrestricted gift. Fellowship recipients are not subject to IP restrictions given the nature of these open research collaboration discussions.<\/p>\n\n\n\n\n\n<p>Funds are being provided as an unrestricted gift to the Institution to be disbursed as per the institution&#8217;s policies and procedures. If you leave the institution, it will be at the discretion of the institution as to whether it will permit and facilitate transferring any funds to your new institution directly. We will not be involved with transfer of funds after the initial payment is complete.<\/p>\n\n\n\n\n\n<p>Fellowship funding is being provided as an unrestricted gift, and standard overhead\/indirect costs would not apply. Since the funding is unrestricted, we will not request nor require detailed accounting of how or when the funds are used, and we do not expect a return of any unused funds.<\/p>\n\n\n\n\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n\n\n<h2 class=\"wp-block-heading\" id=\"2026-microsoft-research-fellows-fellowship-advisors\">2026 Microsoft Research Fellows & Fellowship Advisors<\/h2>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:83%\">\n<hr class=\"wp-block-separator has-text-color has-blue-color has-alpha-channel-opacity has-blue-background-color has-background\" \/>\n\n\n\n<h4 class=\"wp-block-heading has-text-align-left\" id=\"ai-for-global-and-societal-impact-1\">AI for global and societal impact<\/h4>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:16.9%\"><\/div>\n<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Equitable interactive AI tools for creatives<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Melisa-Achoko-Allela_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Melisa Achoko Allela\" class=\"wp-image-1166526\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Melisa-Achoko-Allela_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Melisa-Achoko-Allela_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Melisa-Achoko-Allela_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Melisa-Achoko-Allela_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Melisa Achoko Allela<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>Creative Technologist\/Senior Lecturer, LESO Stories, Namibia University of Science and Technology, Namibia<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/cecilym\/\" id=\"31356\">Cecily Morrison<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kahofman\/\" id=\"32468\">Katja Hofmann<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/baym\/\" id=\"31182\">Nancy Baym<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Global employment, hiring, and performance outcomes based on AI adoption<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Namrata-Kala_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Namrata Kala\" class=\"wp-image-1166528\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Namrata-Kala_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Namrata-Kala_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Namrata-Kala_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Namrata-Kala_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Namrata Kala<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>Associate Professor, Massachusetts Institute of Technology, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/eldillon\/\" id=\"38467\">Eleanor Dillon<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/liraymond\/\" id=\"43509\">Lindsey Raymond<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/horvitz\/\" id=\"32033\">Eric Horvitz<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Grounding AI in global majority languages, cultures, and contexts<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Junjie-Hu_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Junjie Hu\" class=\"wp-image-1166522\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Junjie-Hu_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Junjie-Hu_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Junjie-Hu_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Junjie-Hu_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Junjie Hu<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, University of Wisconsin-Madison, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Binwei-Yao_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Binwei Yao\" class=\"wp-image-1166515\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Binwei-Yao_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Binwei-Yao_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Binwei-Yao_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Binwei-Yao_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Binwei Yao<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, University of Wisconsin-Madison, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Muhammad-Abdel\u2011Mageed_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Muhammad Abdel\u2011Mageed\" class=\"wp-image-1166527\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Muhammad-Abdel\u2011Mageed_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Muhammad-Abdel\u2011Mageed_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Muhammad-Abdel\u2011Mageed_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Muhammad-Abdel\u2011Mageed_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Muhammad Abdel\u2011Mageed<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Associate Professor, The University of British Columbia, Canada<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Sangyun-Kwon_360x360-150x150.jpeg\" alt=\"2026 Microsoft Research Fellowship | Sangyun-Kwon\" class=\"wp-image-1167394\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Sangyun-Kwon_360x360-150x150.jpeg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Sangyun-Kwon_360x360-300x300.jpeg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Sangyun-Kwon_360x360-180x180.jpeg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Sangyun-Kwon_360x360.jpeg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Sangyun Kwan<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, The University of British Columbia, Canada<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Aditya-Vashistha_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Aditya Vashistha\" class=\"wp-image-1166513\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Aditya-Vashistha_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Aditya-Vashistha_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Aditya-Vashistha_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Aditya-Vashistha_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Aditya Vashistha<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, Cornell University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jaoneil\/\" id=\"32172\">Jacki O\u2019Neill<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taganu\/\" id=\"38883\">Tanuja Ganu, <\/a><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xingx\/\" id=\"34906\">Xing Xie<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ogbemie\/\" id=\"43797\">Ogbemi Ekwejunor-Etchie<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Tackling\u202fdisinformation and\u202fmisinformation through the design, implementation and evaluation of provenance tools for end users<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_David-G-Rand_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | David G Rand\" class=\"wp-image-1166518\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_David-G-Rand_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_David-G-Rand_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_David-G-Rand_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_David-G-Rand_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">David Rand<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Professor, Cornell University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Gabrielle-Peloquin-Skulski_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Gabrielle P\u00e9loquin-Skulski\" class=\"wp-image-1166520\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Gabrielle-Peloquin-Skulski_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Gabrielle-Peloquin-Skulski_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Gabrielle-Peloquin-Skulski_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Gabrielle-Peloquin-Skulski_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Gabrielle Peloquin-Skulski<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Massachusetts Institute of Technology, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Dilrukshi-Gamage_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Dilrukshi Gamage\" class=\"wp-image-1166519\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Dilrukshi-Gamage_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Dilrukshi-Gamage_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Dilrukshi-Gamage_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Dilrukshi-Gamage_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Dilrukshi Gamage<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, University of Colombo School of Computing, Sri Lanka<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/asellen\/\" id=\"31112\">Abi Sellen<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/horvitz\/\" id=\"32033\">Eric Horvitz<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:83%\">\n<hr class=\"wp-block-separator has-text-color has-blue-color has-alpha-channel-opacity has-blue-background-color has-background\" \/>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"ai-fundamentals-scalable-reasoning-model-adaption-and-evaluation\">AI fundamentals: scalable reasoning, model adaption and evaluation<\/h4>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:16.9%\"><\/div>\n<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Advancing generative AI evaluation through psychometrics and measurement<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sanmi-Koyejo_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Sanmi Koyejo\" class=\"wp-image-1166532\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sanmi-Koyejo_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sanmi-Koyejo_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sanmi-Koyejo_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sanmi-Koyejo_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Sanmi Koyejo<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, Stanford University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Sang-Truong_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Sang Truong\" class=\"wp-image-1167387\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Sang-Truong_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Sang-Truong_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Sang-Truong_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Sang-Truong_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Sang Truong<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Stanford University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p><\/p>\n\n\n\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wallach\/?msockid=135f78d148036f7b1f416fff49c46e53\">Hanna Wallach<\/a><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xiaoyuanyi\/\" id=\"40768\">, Xiaoyuan Yi<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/mdudik\/\" id=\"32867\">Miro Dudik<\/a>, & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xingx\/\" id=\"34906\">Xing Xie<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Efficient and accurate post-training of retrieval models\u202f<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ravisri-Valluri_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Ravisri Valluri\" class=\"wp-image-1166531\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ravisri-Valluri_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ravisri-Valluri_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ravisri-Valluri_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ravisri-Valluri_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Ravisri Valluri<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, University of California, Los Angeles, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/gauravsinha\/\" id=\"42516\">Gaurav Sinha<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ariy\/\" id=\"36299\">Arun Iyer<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/someh\/\" id=\"37769\">Sonu Mehta<\/a>, Kiran Shiragur & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/shigarg\/\" id=\"44050\">Shivam Garg<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Steering reasoning models with test-time verification<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Hongxiang-Fan_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Hongxiang Fan\" class=\"wp-image-1166521\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Hongxiang-Fan_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Hongxiang-Fan_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Hongxiang-Fan_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Hongxiang-Fan_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Hongxiang Fan<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, Imperial College London<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Mark-Hao-Chen_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Mark (Hao) Chen\" class=\"wp-image-1166523\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Mark-Hao-Chen_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Mark-Hao-Chen_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Mark-Hao-Chen_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Mark-Hao-Chen_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Hao (Mark) Chen<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Imperial College London, United Kingdom<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/amshar\/\" id=\"30997\">Amit Sharma<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nagarajn\/\" id=\"37311\">Nagarajan Natarajan<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/niprasa\/\" id=\"39690\">Niranjani Prasad<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/skarmalkar\/\" id=\"43674\">Sushrut Karmalkar<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:83%\">\n<hr class=\"wp-block-separator has-text-color has-blue-color has-alpha-channel-opacity has-blue-background-color has-background\" \/>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"biological-and-scientific-modeling\">Biological and scientific modeling<\/h4>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:16.9%\"><\/div>\n<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Generative models for regulatory genomics<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sascha-H-Duttke_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Sascha H Duttke\" class=\"wp-image-1166533\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sascha-H-Duttke_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sascha-H-Duttke_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sascha-H-Duttke_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Sascha-H-Duttke_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Sascha Duttke<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, Washington State University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saiman-Dahal_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Saiman Dahal\" class=\"wp-image-1167386\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saiman-Dahal_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saiman-Dahal_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saiman-Dahal_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saiman-Dahal_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Saiman Dahal<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Washington State University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lualex\/\" id=\"41036\">Alex Lu<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kevyan\/\" id=\"39093\">Kevin Yang<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lcrawford\/\" id=\"39660\">Lorin Crawford<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:83%\">\n<hr class=\"wp-block-separator has-text-color has-blue-color has-alpha-channel-opacity has-blue-background-color has-background\" \/>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"foundational-systems-infrastructure-for-ai\">Foundational systems & infrastructure for AI<\/h4>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:16.9%\"><\/div>\n<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">AI for clean electricity planning and deployment<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Maurizio-Porfiri_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Maurizio Porfiri\" class=\"wp-image-1166525\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Maurizio-Porfiri_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Maurizio-Porfiri_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Maurizio-Porfiri_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Maurizio-Porfiri_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Maurizio Porfiri<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Professor, New York University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Camilla-Ancona_240x240-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Camilla Ancona\" class=\"wp-image-1166516\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Camilla-Ancona_240x240-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Camilla-Ancona_240x240-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Camilla-Ancona_240x240.jpg 240w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Camilla Ancona<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>Postdoc, New York University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kstrauss\/\" id=\"32587\">Karin Strauss<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kalytv\/\" id=\"38073\">Kate Lytvynets<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/bnguy\/\" id=\"35942\">Bichlien Nguyen<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jakesmith\/\" id=\"40891\">Jake Smith<\/a> & Danrong Zhang<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Systems, Algorithms, and Rigorous Applications of Reinforcement Learning Post-Training<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xinyu-Lian_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Xinyu Lian\" class=\"wp-image-1166534\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xinyu-Lian_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xinyu-Lian_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xinyu-Lian_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xinyu-Lian_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Xinyu Lian<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, University of Illinois Urbana-Champaign, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/hiballan\/\" id=\"32008\">Hitesh Ballani<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/madanm\/\" id=\"32766\">Madan Musuvathi<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/aashakashah\/\" id=\"43056\">Aashaka Shah<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/abonde\/\" id=\"36068\">Anand Bonde<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ishai\/\" id=\"32116\">Ishai Menache<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kmellou\/\" id=\"38874\">Konstantina Mellou<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/mmolinaro\/\" id=\"42204\">Marco Molinaro<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nswamy\/\" id=\"33138\">Nikhil Swamy<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/rdathathri\/\" id=\"43218\">Roshan Dathathri<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/saikatc\/\" id=\"42411\">Saikat Chakraborty<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sfakhoury\/\" id=\"42180\">Sarah Fakhoury<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sbarke\/\" id=\"43605\">Shraddha Barke<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/siruili\/\" id=\"43857\">Sirui Li<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:83%\">\n<hr class=\"wp-block-separator has-text-color has-blue-color has-alpha-channel-opacity has-blue-background-color has-background\" \/>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"human-ai-collaboration-and-interaction\">Human-AI collaboration and interaction<\/h4>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:16.9%\"><\/div>\n<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">AI Models and Creativity<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Martin-Disley_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Martin Disley\" class=\"wp-image-1166524\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Martin-Disley_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Martin-Disley_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Martin-Disley_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Martin-Disley_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Martin Disley<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, University of Edinburgh, United Kingdom<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/smonroe\/\" id=\"43524\">Shannon Monroe<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/macorwin\/\" id=\"38856\">Matt Corwine<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/rbanks\/\" id=\"33361\">Richard Banks<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/serintel\/\" id=\"33579\">Sean Rintel<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/neberger\/\" id=\"36801\">Neeltje Berger<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">AI-powered Collaborative Dynamic Experiences<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Yining-Rima-Cao_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Yining (Rima) Cao\" class=\"wp-image-1166536\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Yining-Rima-Cao_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Yining-Rima-Cao_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Yining-Rima-Cao_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Yining-Rima-Cao_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Yining (Rima) Cao<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, University of California San Diego, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sianl\/\" id=\"33651\">Si\u00e2n Lindley<\/a>, Nathalie Richie & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/johnwilliams\/\" id=\"40156\">Jack Williams<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Beyond Assistance: AI as teammate for collaboration, alignment, and collective intelligence<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Daniel-Fried_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Daniel Fried\" class=\"wp-image-1166517\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Daniel-Fried_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Daniel-Fried_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Daniel-Fried_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Daniel-Fried_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Daniel Fried<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, Carnegie Mellon University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saujas-Vaduguru_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Saujas Vaduguru\" class=\"wp-image-1167383\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saujas-Vaduguru_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saujas-Vaduguru_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saujas-Vaduguru_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Saujas-Vaduguru_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Saujas Vaduguru<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Carnegie Mellon University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Alan-Ritter_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Alan Ritter\" class=\"wp-image-1166514\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Alan-Ritter_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Alan-Ritter_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Alan-Ritter_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Alan-Ritter_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Alan Ritter<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Associate Professor, Georgia Institute of Technology, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Ethan-Mendes_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Ethan Mendes\" class=\"wp-image-1167393\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Ethan-Mendes_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Ethan-Mendes_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Ethan-Mendes_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2025\/11\/2026-Fellowship_Ethan-Mendes_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Ethan Mendes<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Georgia Institute of Technology, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jenneville\/\" id=\"40946\">Jennifer Neville<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/blog\/tag\/sid-suri\/\" id=\"203811\">Sid Suri<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/kori\/\" id=\"32569\">Kori Inkpen<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sianl\/\" id=\"33651\">Si\u00e2n Lindley<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Toward socially intelligent AI agents<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xuhui-Zhou_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Xuhui Zhou\" class=\"wp-image-1166535\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xuhui-Zhou_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xuhui-Zhou_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xuhui-Zhou_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Xuhui-Zhou_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Xuhui Zhou<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Carnegie Mellon University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jialia\/\" id=\"38470\">Jianxun Lian<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dongsli\/\" id=\"39402\">Dongsheng Li<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/xingx\/\" id=\"34906\">Xing Xie<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/bainguo\/\" id=\"31169\">Baining Guo<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/besh\/\" id=\"42162\">Beibei Shi<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:83%\">\n<hr class=\"wp-block-separator has-text-color has-blue-color has-alpha-channel-opacity has-blue-background-color has-background\" \/>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"multimodal-embodied-intelligence-1\">Multimodal & Embodied Intelligence<\/h4>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:16.9%\"><\/div>\n<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Beyond tokenization: foundational approaches to multimodal LLMs<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Paul-Liang_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Paul Liang\" class=\"wp-image-1166529\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Paul-Liang_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Paul-Liang_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Paul-Liang_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Paul-Liang_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Paul Liang<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, Massachusetts Institute of Technology, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Chanakya-Ekbote_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Chanakya Ekbote\" class=\"wp-image-1167384\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Chanakya-Ekbote_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Chanakya-Ekbote_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Chanakya-Ekbote_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Chanakya-Ekbote_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Chanakya Ekbote<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Massachusetts Institute of Technology, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/vineethn\/\" id=\"44019\">Vineeth N B<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taganu\/\" id=\"38883\">Tanuja Ganu<\/a>, Mercy Ranjit, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/neeraka\/\" id=\"33076\">Neeraj Kayal<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ogbemie\/\" id=\"43797\">Ogbemi Ekwejunor-Etchie<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">General-purpose robotic foundation models for embodied AI<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Jiajun_Wu_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Jiajun Wu\" class=\"wp-image-1167385\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Jiajun_Wu_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Jiajun_Wu_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Jiajun_Wu_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/04\/2026-Fellowship_Jiajun_Wu_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Jiajun Wu<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Assistant Professor, Stanford University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziwei-Liu_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Ziwei Liu\" class=\"wp-image-1166512\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziwei-Liu_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziwei-Liu_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziwei-Liu_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziwei-Liu_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Ziwei Liu<\/h5>\n\n\n\n<p><strong>Fellowship Advisor<\/strong><br>Professor, Nanyang Technological University, Singapore<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziqi-Huang_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Ziqi Huang\" class=\"wp-image-1166511\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziqi-Huang_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziqi-Huang_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziqi-Huang_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Ziqi-Huang_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Ziqi Huang<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Nanyang Technological University, Singapore<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jiaoyan\/\" id=\"36125\">Jiaolong Yang<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lizo\/\" id=\"36152\">Li Zhao<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lisu\/\" id=\"32703\">Lily Sun<\/a>, <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jfgao\/\" id=\"32246\">Jianfeng Gao<\/a> & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/bainguo\/\" id=\"31169\">Baining Guo<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n\n<div style=\"padding-bottom:0; padding-top:0\" class=\"wp-block-msr-immersive-section alignfull row wp-block-msr-immersive-section\">\n\t\n\t<div class=\"container\">\n\t\t<div class=\"wp-block-msr-immersive-section__wrapper\">\n\t\t\t<h5 class=\"wp-block-heading is-style-default\" id=\"theme-name\">Robot learning from alternative data sources\u202f<\/h5>\n\n\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:33%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:30%\">\n<figure class=\"wp-block-image size-thumbnail is-style-rounded\"><img loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" src=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Rao_Fu_360x360-150x150.jpg\" alt=\"2026 Microsoft Research Fellowship | Rao Fu\" class=\"wp-image-1166530\" style=\"object-fit:cover\" srcset=\"https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Rao_Fu_360x360-150x150.jpg 150w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Rao_Fu_360x360-300x300.jpg 300w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Rao_Fu_360x360-180x180.jpg 180w, https:\/\/www.microsoft.com\/en-us\/research\/wp-content\/uploads\/2026\/03\/2026-Fellowship_Rao_Fu_360x360.jpg 360w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/figure>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:70%\">\n<h5 class=\"wp-block-heading\" id=\"firstname-lastname\">Rao Fu<\/h5>\n\n\n\n<p><strong>Fellow<\/strong><br>PhD Student, Brown University, United States<\/p>\n\n\n\n<div style=\"height:5px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-9d6595d7 wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:66.66%\">\n<p>Microsoft Research lead(s): Michael Murray, Tess Hellebrekers & <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tanreuben\/\" id=\"43827\">Reuben Tan<\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:17%\"><\/div>\n<\/div>\t\t<\/div>\n\t<\/div>\n\n\t<\/div>\n\n\n","protected":false},"featured_media":1154559,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_hide_image_in_river":null,"footnotes":""},"msr-opportunity-type":[155533],"msr-region":[256048],"msr-locale":[268875],"msr-program-audience":[],"msr-post-option":[269148,269142],"msr-impact-theme":[],"class_list":["post-1150350","msr-academic-program","type-msr-academic-program","status-publish","has-post-thumbnail","hentry","msr-opportunity-type-grants-and-fellowships","msr-region-global","msr-locale-en_us","msr-post-option-approved-for-river","msr-post-option-include-in-river"],"msr_description":"","msr_social_media":[],"related-researchers":[],"tab-content":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-academic-program\/1150350","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-academic-program"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-academic-program"}],"version-history":[{"count":184,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-academic-program\/1150350\/revisions"}],"predecessor-version":[{"id":1168290,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-academic-program\/1150350\/revisions\/1168290"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/1154559"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1150350"}],"wp:term":[{"taxonomy":"msr-opportunity-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-opportunity-type?post=1150350"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=1150350"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1150350"},{"taxonomy":"msr-program-audience","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-program-audience?post=1150350"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1150350"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1150350"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}