13 stories tagged llm-reasoning, newest first.
41 points by porridgeraisin | 21 comments
Special topic article by Subbarao Kambhampati in AI Magazine on how LLMs, as System 1 predictors, can be made to exhibit System 2 deliberative reasoning through post-training.
491 points by ModelForge | 158 comments
Sebastian Raschka's Ahead of AI newsletter post examining OpenAI's GPT-6 Astra, looped transformers with recurrent depth, and hidden chain-of-thought reasoning research.
1353 points by novemp | 837 comments
A Mastodon post presenting the riddle-like question of whether to walk or drive to a car wash 50 meters away, likely highlighting LLM reasoning failures.
429 points by meetpateltech | 215 comments
Google blog post announcing the rollout of Gemini 2.5 Deep Think in the Gemini app, with mathematicians getting access to the IMO competition version.
444 points by sxv | 228 comments
Science magazine's ScienceAdviser newsletter covering how irrelevant cat facts confuse LLMs on math problems, plus other science news items.
440 points by Davidzheng | 654 comments
X post by OpenAI researcher Alexander Wei announcing an experimental reasoning LLM achieved gold medal-level performance at the 2025 International Math Olympiad.
376 points by trott | 126 comments
arXiv paper presenting LIMO, a model showing complex math reasoning in LLMs can be elicited with only 817 curated training samples.
527 points by mfiguiere | 374 comments
OpenAI announcement of deep research, a ChatGPT agent that uses reasoning to synthesize online information and complete multi-step research tasks.
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.
529 points by optimalsolver | 490 comments
OpenReview paper page reporting that OpenAI's o1-preview loses about 30% accuracy on slightly varied Putnam math competition problems.
334 points by TaurenHunter | 138 comments
An AI Papers Academy article and video explaining Meta AI's Coconut paper on LLM reasoning via chain of continuous thought in latent space.
626 points by dselsam | 404 comments
Mastodon post by mathematician Terence Tao describing his experiments with OpenAI's o1 model on research-level mathematical tasks.
641 points by loganfrederick | 566 comments
Simon Willison's blog post with notes on OpenAI's o1-preview and o1-mini models, covering chain-of-thought training, API details, hidden reasoning tokens and examples.