#reinforcement learning
Reinforcement Learning: 35 AI articles covering reinforcement learning news, analysis, and research
Articles
Reinforcement Learning Cuts Routing Violations in Dense Chip⭐9
NYU researchers apply reinforcement learning to detailed routing, significantly reducing design rule violations in dense chip layouts at advanced nodes.
Google Research Unveils Retrieve-for-Train (R4T): An RL-Compiled⭐8
Google Research's Retrieve-for-Train (R4T) uses RL and a diffusion retriever to generate diverse query fan-outs 12–20× faster, avoiding paraphrastic collapse in...
OpenAI Releases Model Misalignment Disclosure Framework With 3⭐10
OpenAI releases a misalignment disclosure framework with three review tracks and six incident reports from RL training, tightening transparency for frontier AI ...
Cognition Releases SWE-2: A Kimi K3 Post-Trained Coding Model⭐7
Cognition's SWE-2 coding model, post-trained from Kimi K3, hits 50.0% on FrontierCode 1.1, matching Fable 5.1 at 64% lower cost and running only inside Devin.
Dynamical System Transfer Learning with Reduced Order Models⭐9
Reduced order models accelerate transfer learning in RL for dynamical systems, cutting training costs while preserving key physics for faster policy adaptation.
Training a Coding Model to Paint Watercolours with TRL and OpenEnv⭐9
Training a coding model to paint watercolours: explore how TRL and OpenEnv use reinforcement learning to turn code generation into AI-driven artistry.
Google AI Introduces EnvHarness: A Programmable Layer Turning⭐7
Google's EnvHarness adapts agent training environments dynamically, boosting performance by up to 9 points with 9.8% fewer steps.
Hugging Face Introduces Microduck: A $399 Open-Source 25 cm⭐7
Microduck is a $399 open-source 25 cm biped robot from Hugging Face, trained via reinforcement learning for walking, kicking, and more.
Barret Zoph, Thinking Machines Co-Founder Who Returned to OpenAI⭐7
AI executive Barret Zoph rejoins Google as VP of Research after brief OpenAI return, bringing RL expertise to Gemini.
Hugging Face Launches Cute, Open-Source Microduck Robot for $399⭐8
Hugging Face unveils Microduck, a $399 open-source robot with reinforcement learning, camera, lidar, and object pickup. Ships for Christmas.