Research Focus: Week of November 22, 2023
A new deep-learning compiler for dynamic sparsity; Tongue Tap could make tongue gestures viable for VR/AR headsets; Ranking LLM-Generated Loop Invariants for Program Verification; Assessing the limits of zero-shot foundation models in single-cell biology.
In the news | TheSequence
The Sequence Chat: Doug Burger- Technical Fellow, Microsoft Research About Building Autonomous Agents, AutoGen and the Future of Generative AI
One of the collaborators in the AutoGen project, shares insights about its vision, architecture and the future of autonomous agents.
In the news | IEEE
Karin Strauss, 2024 IEEE Fellow
Congratulations to Karin Strauss, Senior Principal Research Manager in AI4Science, who was recently named as a 2024 IEEE Fellow!
刷新自我，引领计算的新范式
作者：周礼栋 在2023年即将远去的这一刻，我不禁会产生这样一种遐想：若干年后，未来的人们会回想起这一年中大家兴奋地学习使用大语言模型（LLMs）和各种 Copilot 的情景，并感慨道：“改变就是从那时开始发生的”。尽管2023年的我们对人工智能的了解还在初级阶段，但已确信它将与人类的命运紧密相连。 对微软亚洲研究院而言，2023年则更多了一些新的里程碑意义：我们迎来了建院25周年。回首过去的2...
In the news | Possible
Peter Lee on the Future of Health and Medicine
A conversation with Peter Lee on the POSSIBLE Podcast. "At some point, it just makes sense that, of course, you would never practice medicine without washing your hands first. And we will definitively get to a point—probably sooner than we…
In the news | Communications of the ACM
What Would the Chatbot Say?
Sebastien Bubeck is a senior principal research manager in the Machine Learning Foundations Group at Microsoft Research. Bubeck often generates stories about unicorns for his young daughter using a chatbot powered by GPT-4, the latest large language model (LLM) by…
In the news | Scientific American
When It Comes to AI Models, Bigger Isn’t Always Better
Artificial intelligence models are getting bigger, along with the data sets used to train them. But scaling down could solve some big AI problems. Artificial intelligence has been growing in size. The large language models (LLMs) that power prominent chatbots,…
Orca 2: Teaching Small Language Models How to Reason
| Ahmed Awadallah, Andres Codas, Luciano Del Corro, Hamed Khanpour, Shweti Mahajan, Arindam Mitra, Hamid Palangi, Corby Rosset, Clarisse Simoes Ribeiro, and Guoqing Zheng
At Microsoft, we’re expanding AI capabilities by training small language models to achieve the kind of enhanced reasoning and comprehension typically found only in much larger models.
Lifelong model editing in large language models: Balancing low-cost targeted edits and catastrophic forgetting
| Tom Hartvigsen and Hamid Palangi
Lifelong model editing fixes mistakes discovered after model deployment. This work could expand sequential editing to model properties like fairness and privacy and enable a new class of solutions for adapting LLMs over long deployment lifetimes.