News & features
Orca-AgentInstruct: Agentic flows can be effective synthetic-data generators
| Arindam Mitra, Ahmed Awadallah, and Yash Lara
Orca-AgentInstruct, from Microsoft Research, can generate diverse, high-quality synthetic data at scale to post-train and fine-tune base LLMs for expanded capabilities, continual learning, and increased performance.
Research Focus: Week of October 28, 2024
New Research | FLASH: Workflow automation agent for diagnosing recurring incidents; METAREFLECTION: Learning instructions for language agents using past reflections; Boosting LLM training efficiency through faster communication between GPUs; and more.
Orca-Math: Demonstrating the potential of SLMs with model specialization
| Arindam Mitra, Hamed Khanpour, Corby Rosset, and Ahmed Awadallah
Microsoft’s Orca-Math, a specialized small language model, outperforms much larger models in solving math problems that require multi-step reasoning and shows the potential of using feedback to improve language models. Learn more.
AI Frontiers: The future of scale with Ahmed Awadallah and Ashley Llorens
| Ahmed Awadallah and Ashley Llorens
What’s the driving force behind AI’s recent, rapid progress? Research manager Ahmed Awadallah shares his insights on this, the two-stage approach to training large-scale models, and the need for better model evaluation in this episode of the #MSRPodcast.
AI Explainer: Foundation models and the next era of AI
| Ahmed Awadallah
The release of OpenAI’s GPT-4 is a significant advance that builds on several years of rapid innovation in foundation models. GPT-4, which was trained on the Microsoft Azure AI supercomputer, has exhibited significantly improved abilities across many dimensions—from summarizing lengthy…
You get what you measure: New NLU benchmarks for few-shot learning and robustness evaluation
| Jianfeng Gao and Ahmed Awadallah
Recent progress in natural language understanding (NLU) has been driven in part by the availability of large-scale benchmarks that provide an environment for researchers to test and measure the performance of AI models. Most of these benchmarks are designed for…
Conversations with data: Advancing the state of the art in language-driven data exploration
| Alex Polozov, Chris Meek, and Ahmed Awadallah
One key aspiration of AI is to develop natural and effective task-oriented conversational systems. Task-oriented conversational systems use a natural language interface to collaborate with and support people in accomplishing specific goals and activities. They go beyond chitchat conversation. For…
Awards | British Computer Society
Ahmed Awadallah awarded the 2020 Karen Sparck Jones Award
Ahmed Awadallah received the 2020 Karen Sparck Jones Award presented by the British Computer Society (BCS). This award is given to researchers within 10 years of their PhD who have contributed significantly to information retrieval and/or natural language processing.
In the news | TheNextWeb
Microsoft’s new AI can generate smart to-do lists from your emails
Researchers from the University of Washington and Microsoft’s AI team today unveiled a ‘Smart To-Do’ tool for automatically generating task lists from emails. Smart To-Do is an AI feature that scans your outgoing emails for actionable text and turns your…