Our new piece in @TheLancet on the potential of AI agents in medicine @james_y_zou https://
thelancet.com/journals/lance
t/article/PIIS0140-6736(25)00202-8/fulltext
…
RESEARCH
-

AI Agents Transform Medical Practice and Patient Care
By
–
-
AI Execution vs Ideas: New Podcast Insights
By
–
"with AI, execution won't matter, only ideas" — check out the podcast with @amasad & @mattturck @FirstMarkCap https://t.co/Bu1FXLSfPp
— Replit ⠕ (@Replit) 6 février 2025"with AI, execution won't matter, only ideas" — check out the podcast with @amasad & @mattturck @FirstMarkCap
-

How LLMs Learn to Reason: Comprehensive Study Reveals
By
–
Trending on alphaXiv (2/6): The most comprehensive (and refreshingly clear) study on how LLMs learn to reason. A systematic investigation reveals key ingredients: SFT initialization helps, reward shaping stabilizes training, filtered verifiable rewards improve generalization,
-

Stanford AI Projects Improve Clinical Reports and Hearing Assessments
By
–
For the first time, @StanfordHAI
’s Seed Research Grants program is co-funding two projects with @StanfordCDH
: one to improve patient experience by summarizing clinical reports, and another to enable accessible, accurate, at-home hearing assessments. https://
stanford.io/42OK3nl -
Industrial Robot Sorting Error Decreases With Training Data
By
–
How industrial robot sorting error decreases as training data is added (from @AmbiRobotics
): -
DeepSeek-R1 RL Training Simplicity Versus Process Reward Models
By
–
From an open-research point of view, probably the greatest thing about DeepSeek–R1 is how its RL training technique appears so straightforward and simple in comparison to the cumbersome approaches people were starting to think necessary for learning reasoning like Process Reward
-

GSK Tackles AI Hallucinations Problem in Drug Development
By
–
Hallucinations in #AI: How GSK is addressing a critical problem in drug development
by @Bryson_M @VentureBeat Learn more: https://
buff.ly/42mIHjk #ArtificialIntelligence #MI #MachineLearning #Tech #Technology cc: @amuellerml @sallyeaves @marcusborba -

Evolution of RAG Architectures: From Naive to Multi-Agent Systems
By
–
Retrieval-Augmented Generation (RAG) is evolving fast! This infographic breaks down different RAG architectures – from Naïve RAG to Multi-Agent RAG – enhancing retrieval, ranking, and generation for AI applications. What’s the future of RAG? Multi-modal, agentic, and
-
Schmidhuber’s 1991 RNN History Compression Method for Long Context
By
–
PS @SchmidhuberAI in 1991 published something related to compressing history length in RNNs, which may be very relevant for long context, an important topic of research today.