{"id":1135033,"date":"2025-03-26T09:00:00","date_gmt":"2025-03-26T16:00:00","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?p=1135033"},"modified":"2025-04-07T17:43:03","modified_gmt":"2025-04-08T00:43:03","slug":"research-focus-week-of-march-24-2025","status":"publish","type":"post","link":"https:\/\/www.microsoft.com\/en-us\/research\/blog\/research-focus-week-of-march-24-2025\/","title":{"rendered":"Research Focus: Week of March 24, 2025"},"content":{"rendered":"\n

In this issue:<\/strong><\/p>\n\n\n\n

We examine a new conversation segmentation method that delivers more coherent and personalized agent conversation, and we review efforts to improve MLLMs\u2019 understanding of geologic maps. Check out the latest research and other updates.<\/p>\n\n\n\n

\"Research<\/figure>\n\n\n\n
\n

NEW RESEARCH<\/h2>\n\n\n\n

SeCom: On Memory Construction and Retrieval for Personalized Conversational Agents<\/h3>\n\n\n\n

Researchers from Microsoft and Tsinghua University propose a new method to help conversational AI agents deliver more coherent and personalized responses during complex long-term dialogue.<\/p>\n\n\n\n

Large language models (LLMs) are widely used to enable more complicated discussions across a broader range of topics than traditional dialogue systems. However, managing excessively long context that contains irrelevant information is a major challenge. Existing solutions typically perform retrieval augmented response generation by constructing memory banks from conversation history at either the turn-level, session-level, or through summarization.<\/p>\n\n\n\n

The proposed new approach, SeCom<\/a>, constructs the memory bank at segment level by introducing a conversation Se<\/strong>gmentation model that partitions long-term conversations into topically coherent segments, while applying Com<\/strong>pression based denoising on memory units to enhance memory retrieval. Experimental results show that SeCom<\/strong> exhibits a significant performance advantage over baselines on long-term conversation benchmarks LOCOMO and Long-MT-Bench+. Additionally, the proposed conversation segmentation method demonstrates superior performance on dialogue segmentation datasets such as DialSeg711, TIAGE, and SuperDialSeg. <\/p>\n\n\n\n

\n
Read the paper<\/a><\/div>\n<\/div>\n\n\n\n
\n<\/div>\n\n\n\n
\n

NEW RESEARCH<\/h2>\n\n\n\n

PEACE: Empowering Geologic Map Holistic Understanding with MLLMs<\/h3>\n\n\n\n

Microsoft Researchers and external colleagues introduce GeoMap-Agent<\/a>, an AI system specifically designed for geologic map understanding and analysis. In the lab, they measure its effectiveness using a new benchmark called GeoMap-Bench, a novel gauge for evaluating multimodal large language models (MLLMs) in geologic map understanding. Geologic maps provide critical insights into the structure and composition of Earth’s surface and subsurface. They are indispensable in fields including disaster detection, resource exploration, and civil engineering.<\/p>\n\n\n\n

Current MLLMs often fall short in understanding geologic maps, largely due to the challenging nature of cartographic generalization, which involves handling high-resolution maps, managing multiple associated components, and requiring domain-specific knowledge.<\/p>\n\n\n\n

This paper presents results of experiments in which GeoMap-Agent achieves an overall score of 0.811 on GeoMap-Bench, significantly outperforming the 0.369 score of GPT-4o. The researchers intend to enable advanced AI applications in geology, powering more efficient and accurate geological investigations.<\/p>\n\n\n\n

\n
Read the paper<\/a><\/div>\n<\/div>\n\n\n\n
\n<\/div>\n\n\n\n
\n

NEW RESEARCH<\/h2>\n\n\n\n

The future of the industrial AI edge is cellular<\/h3>\n\n\n\n

Reliable, high-bandwidth wireless connectivity and local processing at the edge are crucial enablers for emerging industrial AI applications. This work proposes that cellular networking is the ideal connectivity solution for these applications, due to its virtualization and support for open APIs. The researchers project the emergence of a converged industrial AI edge encompassing both computing and connectivity, in which application developers leverage the API to implement advanced functionalities. They present a case study showing evidence of the effectiveness of this approach, evaluated on an enterprise-grade 5G testbed.<\/p>\n\n\n\n

\n
Read the paper<\/a><\/div>\n<\/div>\n\n\n\n
\n<\/div>\n\n\n\n
\n

NEW RESEARCH<\/h2>\n\n\n\n

RE#: High Performance Derivative-Based Regex Matching with Intersection, Complement, and Restricted Lookarounds<\/h3>\n\n\n\n

A regular expression (regex or RE) is a sequence of characters used to match, search, and manipulate strings in text based on specific criteria. REs are used in programming languages for data validation, text parsing, and search operations.<\/p>\n\n\n\n

This paper<\/a> presents a tool and theory built on\u202fsymbolic derivatives that does not use backtracking, while supporting both classical operators and complement, intersection, and restricted lookarounds. The researchers show that the main matching algorithm has\u202finput-linear\u202fcomplexity both in theory as well as experimentally. They apply thorough evaluation on popular benchmarks that show that RE# is over 71% faster than the next fastest regex engine in Rust on the baseline, and\u202foutperforms all state-of-the-art engines on extensions of the benchmarks, often by several orders of magnitude. <\/p>\n\n\n\n

This work could potentially enable new applications in LLM prompt engineering frameworks, new applications in medical research and bioinformatics, and new opportunities in access and resource policy language design by web service providers.<\/p>\n\n\n\n

\n
Read the paper<\/a><\/div>\n<\/div>\n\n\n\n
\n<\/div>\n\n\n\n
\n

NEW RESEARCH<\/h2>\n\n\n\n

Toward deep learning sequence\u2013structure co-generation for protein design<\/h3>\n\n\n\n

Researchers review recent advances in deep generative models for protein design, with a focus on sequence-structure co-generation methods. They describe the key methodological and evaluation principles underlying these methods, highlight recent advances from the literature, and discuss opportunities for continued development of sequence-structure co-generation approaches.<\/p>\n\n\n\n

Deep generative models that learn from the distribution of natural protein sequences and structures may enable the design of new proteins with valuable functions. While most of today\u2019s models focus on generating either sequences or structures, emerging co-generation methods promise more accurate and controllable protein design, ideally achieved by modeling both modalities simultaneously. <\/p>\n\n\n\n

\n
Read the paper<\/a><\/div>\n<\/div>\n\n\n\n\t
\n\t\t\n\n\t\t

\n\t\tPODCAST SERIES<\/span>\n\t<\/p>\n\t\n\t

\n\t\t\t\t\t\t
\n\t\t\t\t\n\t\t\t\t\t\"Illustrated\n\t\t\t\t<\/a>\n\t\t\t<\/div>\n\t\t\t\n\t\t\t
\n\n\t\t\t\t\t\t\t\t\t

The AI Revolution in Medicine, Revisited<\/h2>\n\t\t\t\t\n\t\t\t\t\t\t\t\t

Join Microsoft\u2019s Peter Lee on a journey to discover how AI is impacting healthcare and what it means for the future of medicine.<\/p>\n\t\t\t\t\n\t\t\t\t\t\t\t\t

\n\t\t\t\t\t
\n\t\t\t\t\t\t\n\t\t\t\t\t\t\tListen now\t\t\t\t\t\t<\/a>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t<\/div>\n\t<\/div>\n\t<\/div>\n\n\n\n
\n

PODCAST<\/h2>\n\n\n\n

New Series: The AI Revolution in Medicine, Revisited<\/h3>\n\n\n\n

Two years ago, OpenAI\u2019s GPT-4 kick-started a new era in AI. In the months leading up to its public release, Peter Lee<\/a>, president of Microsoft Research, cowrote The AI Revolution in Medicine: GPT-4 and Beyond<\/em> (opens in new tab)<\/span><\/a>, a book full of optimism for the potential of advanced AI models to transform the world of healthcare. In this special Microsoft Research Podcast<\/a> series, Lee revisits the book, exploring how patients, providers, and other medical professionals are experiencing and using generative AI today while examining what he and his coauthors got right\u2014and what they didn\u2019t foresee.<\/p>\n\n\n\n

\n