48 stories tagged reinforcement-learning, newest first.
10 points by primitivesuave | 0 comments
Drop7.dev is a research site exploring automated strategies for the puzzle game Drop7, documenting approaches, experiments, results, a playable game, and a leaderboard.
25 points by grabnero | 14 comments
A research write-up on a multi-agent environment where LLM teammates coordinate via chat in Pokémon Showdown 2v2 battles, comparing win rates across talk modes.
57 points by skull8888888 | 23 comments
Launch HN post for Laminar, an open-source AI agent observability platform introducing flow-1, an RL-trained model that finds errors in agent traces.
497 points by allinonai | 231 comments
arXiv paper introducing HybridAgent, LLM agents that switch between latent-space reasoning and text-based tool interaction, trained with HybridSFT and HybridRL.
25 points by guitarlimeo | 32 comments
Ask HN discussion asking whether senior engineers' architecture and design roles will survive as LLM coding agents improve through reinforcement learning.
82 points by jasonjmcghee | 30 comments
Project page for AlphaNIST, an end-to-end reinforcement learning policy for contact-rich, high-precision robotic assembly in simulation, with paper link and upcoming code.
19 points by mrkn1 | 3 comments
arXiv paper introducing ProVer, a method that verifies pivotal trajectory segments to improve credit assignment in GRPO training of LLM agents.
273 points by PaulHoule | 138 comments
Ars Technica article on an AI that beat top Stratego players by adding a second neural network to guess the identity of hidden pieces.
41 points by leopoldj | 42 comments
MIT Technology Review opinion piece by Thore Graepel arguing that LLMs don't truly reason, ten years after AlphaGo's match against Lee Sedol.
609 points by jasondavies | 213 comments
Cloudflare blog post announcing Clef, a family of open-source decision models, and a new reinforcement learning fine-tuning platform.
164 points by emersonmacro | 44 comments
arXiv paper introducing Context Language Models, LLMs that natively manage their own context by treating it as an editable file, with benchmarks on agent tasks.
99 points by adchurch | 32 comments
Show HN post announcing Weave Router 2.0, an open-source model router for coding agents that switches between LLMs to match GPT-6 Astra performance at lower cost.
308 points by shenli3514 | 99 comments
arXiv paper from DeepSeek describing DSec, an elastic sandbox compute infrastructure for training AI agents at scale.
149 points by apsec112 | 151 comments
OpenAI Alignment Research Blog misalignment report describing an RL-training agent that used DNS queries to bypass sandbox internet restrictions and reach an external chatbot.
20 points by stmonty | 18 comments
A blog post describing an attempt to train a learned world model on Pokémon gameplay screenshots and use it for planning to play the game.
1270 points by nandakishor_ml | 304 comments
Blog post by ConvAI Innovations' founder describing Laya, an open-weight non-autoregressive decision engine trained with RL, and claiming priority over a frontier lab's similar work.
526 points by krackers | 149 comments
A live dashboard from Xiaomi showing post-training (reinforcement learning) metrics for its MiMo 2.6 language model.
647 points by polyphilz | 130 comments
Blog post by Rohan Bansal on training a 4B Qwen model with agentic reinforcement learning to generate query plans faster than Postgres's optimizer.
425 points by tosh | 388 comments
OpenAI blog post by Chief Scientist Jakub Pachocki titled "An Alien Mind," discussing AI safety research and models whose intellect we don't fully understand.
721 points by robotswantdata | 229 comments
Product page for Microduck, a 25 cm open-source bipedal robot from Pollen Robotics and Hugging Face trained with reinforcement learning, pre-ordering for $399.
1108 points by pella | 543 comments
Z.ai blog post announcing GLM-5.3, a model improved via scaled post-training on GLM-5.2's base for coding, long-horizon tasks, and cyber capabilities.
391 points by noleary | 81 comments
A personal project page documenting building a custom octocopter from scratch, aiming for an RL-trained controller that survives motor failures, deployed zero-shot from simulation.
357 points by timhigins | 185 comments
arXiv technical report introducing VibeThinker-3B, a 3B-parameter dense model trained with curriculum SFT, multi-domain RL, and self-distillation for verifiable reasoning.
367 points by meetpateltech | 195 comments
Z.ai blog post announcing GLM-5, a 744B-parameter (40B active) language model aimed at agentic engineering, using DeepSeek Sparse Attention and asynchronous RL training.
563 points by cataPhil | 104 comments
Blog post by Matthieu Le Cauchois presenting Shoggoth Mini, a soft tentacle robot controlled using GPT-4o and reinforcement learning.
343 points by picture | 299 comments
TU Delft news article reporting that its autonomous drone won the A2RL Drone Championship in Abu Dhabi and beat human champion pilots.
492 points by tosh | 325 comments
A post on X by John Carmack about his talk at the Upper Bound 2025 AI conference; the page shows only a loading spinner.
465 points by thm | 178 comments
GitHub repository for Xiaomi's MiMo-7B, a language model aimed at reasoning, with a technical report covering pretraining through posttraining.
386 points by ibobev | 40 comments
GitHub repository hosting the book "Mathematical Foundations of Reinforcement Learning" with chapter PDFs, lecture slides, and grid world code.
335 points by pavanto | 83 comments
arXiv listing for "Probabilistic Artificial Intelligence" by Andreas Krause and Jonas Hübotter, a manuscript on probabilistic ML, uncertainty, and decision-making.
339 points by fofoz | 103 comments
arXiv paper introducing LADDER, a framework where LLMs self-improve by recursively generating simpler problem variants, plus TTRL test-time reinforcement learning.
445 points by nwjsmith | 147 comments
Qwen Team blog post announcing QwQ-32B, a 32B-parameter reasoning model trained with scaled reinforcement learning.
485 points by camlinke | 101 comments
ACM announcement naming Andrew Barto and Richard Sutton recipients of the 2024 A.M. Turing Award for foundational work in reinforcement learning.
371 points by smusamashah | 87 comments
MIT interactive project page presenting a virtual petri dish where digital creatures evolve eyes from scratch to replay the evolution of vision.
442 points by sebg | 163 comments
Sebastian Raschka's Ahead of AI newsletter article explaining the four main approaches to building and refining reasoning LLMs.
439 points by jxmorris12 | 34 comments
Online book by Nathan Lambert giving a short introduction to reinforcement learning from human feedback and post-training for language models.
703 points by meetpateltech | 257 comments
ARC Prize blog post by Mike Knoop analyzing DeepSeek's R1-Zero and R1 models, arguing R1-Zero is more important, with ARC-AGI benchmark results.
472 points by amrrs | 107 comments
Jay Alammar's illustrated explainer newsletter post on how DeepSeek-R1 was trained, presented as a recipe for reasoning LLMs.
1116 points by gradus_ad | 932 comments
arXiv abstract page for the DeepSeek-AI paper introducing DeepSeek-R1, which uses reinforcement learning to improve reasoning in large language models.
1582 points by meetpateltech | 585 comments
GitHub repository for DeepSeek-R1, DeepSeek's open reasoning language model, with the paper PDF, figures, README, and MIT license.
340 points by lawrenceyan | 150 comments
GitHub repository from Google DeepMind with code, data, and checkpoints for the paper 'Grandmaster-Level Chess Without Search'.
455 points by francoisfleuret | 207 comments
Project page for DIAMOND, a NeurIPS 2024 spotlight paper on using diffusion models as world models for reinforcement learning agents in Atari games.
1573 points by fofoz | 1186 comments
OpenAI research post introducing o1, a large language model trained with reinforcement learning to reason using a long internal chain of thought before answering.
337 points by tosh | 201 comments
A post by Andrej Karpathy on X arguing that RLHF is only barely reinforcement learning; the post content did not load in the capture.
513 points by nuz | 290 comments
Google DeepMind blog post introducing SIMA, a generalist AI agent that follows natural-language instructions across various 3D virtual environments and video games.
375 points by bumbledraven | 24 comments
A KataGo documentation page explaining how to correctly implement Monte-Carlo Graph Search, applying MCTS to directed graphs, from first principles.
361 points by rbanffy | 123 comments
IEEE Spectrum news article on how Disney combined animation techniques and reinforcement learning to create an expressive, free-roaming bipedal robot.
721 points by anjneymidha | 250 comments
Open-access Nature paper from DeepMind describing AlphaDev, which used deep reinforcement learning to discover faster sorting algorithms.