Microsoft Research introduces Agent Lightning to connect AI agents to RL training
Microsoft Research introduced Agent Lightning, a tool that connects existing AI agents to reinforcement learning training without requiring them to be rebuilt.
#microsoft-research
View all AI newsMicrosoft Research introduces Quine, an early-stage multimodal world model of biology
Microsoft Research describes Quine as an early-stage effort to connect biological scales and modalities so scientists can computationally search and prioritize hypotheses before laboratory work.
Agensh coordinates up to 1,024 coding agents without a central orchestrator
A Microsoft Research paper describes Agensh, a self-organized harness that coordinates parallel coding agents through shared state. With GPT-5.6-sol, scaling from 1 to 128 agents raises the mean final test-pass rate on the five hardest ProgramBench tasks from 19.31% to 28.78%; on pandoc, 1,024 agents raise it from 33.89% to 55.06%.
Microsoft Research findings show moving AI inference beyond the robot can improve task success
The findings also say the approach can increase efficiency and support more advanced physical AI workloads.
AutoTailor filters auto-generated web APIs from 1,283 candidates to 33
DAIR.AI highlighted AutoTailor, a paper from Microsoft Research and colleagues that turns web trajectories into parameterized browser-automation APIs and filters them down to 33. On 106 WebArena Postmill tasks, those APIs with a ReAct fallback reached 90.6% correctness versus 87.5% for ReAct alone.
Microsoft Research: Nature paper highlights RetroChimera for chemical synthesis
Microsoft Research said a new Nature paper highlights RetroChimera, a predictive model that helps accelerate chemical synthesis so researchers can explore a wide range of molecules.