GAIR-NLP

O1 Replication Journey

AlphaGo Moment for Model Architecture Discovery.

[COLM 2025] LIMO: Less is More for Reasoning

FacTool: Factuality Detection in Generative AI

[Extended verision ICLR 2025 Blog Track] Anole: An Open, Autoregressive and Native Multimodal Models for Interleaved Image-Text Generation

Scaling Deep Research via Reinforcement Learning in Real-world Environments.

OpenResearcher, an advanced Scientific Research Assistant

[NeurlPS D&B 2024] Generative AI for Math: MathPile

SOTA Math Opensource LLM

PC Agent: While You Sleep, AI Works - A Cognitive Journey into Digital World

[ICML 2025] Programming Every Example: Lifting Pre-training Data Quality Like Experts at Scale

Generative Judge for Evaluating Alignment

Generative AI Act II: Test Time Scaling Drives Cognition Engineering

Revisiting Mid-training in the Era of Reinforcement Learning Scaling

LIMI: Less is More for Agency

Rethinking RL Scaling for Vision Language Models: A Transparent, From-Scratch Framework and Comprehensive Evaluation Scheme

[ICLR 2026] Efficient Agent Training for Computer Use

[COLM 2026] MegaScience: Pushing the Frontiers of Post-Training Datasets for Science Reasoning

Doodling our way to AGI ✏️ 🖼️ 🧠

Reformatted Alignment

VidaForge: Open Research Infrastructure for Video Pretraining Data Recipes

[NeurIPS 2024] OlympicArena: Benchmarking Multi-discipline Cognitive Reasoning for Superintelligent AI

[ACL2026 Main] AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts

[ICML 2026 Oral] Agent-native Mid-training for Software Engineering

Official implementation for 'Extending LLMs’ Context Window with 100 Samples'

[AAAI 2025 oral] Evaluating Mathematical Reasoning Beyond Accuracy

Benchmarking Benchmark Leakage in Large Language Models

[ICLR 2026] SR-Scientist: Scientific Equation Discovery With Agentic AI

ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry