Stars
Can AI agents rebuild websites through browser exploration?
OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks
OpenResearcher: A Fully Open Pipeline for Long-Horizon Deep Research Trajectory Synthesis
The code and data for "MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark" [NeurIPS 2024]
BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent (ACL 2026 Main)
A version of verl to support diverse tool use [TMLR 2026]
Enable Comprehensive LLM Evaluation on Graph Reasoning
Official repository for paper "ReasonIR Training Retrievers for Reasoning Tasks".
Unify Efficient Fine-tuning of RAG Retrieval, including Embedding, ColBERT, ReRanker.
General Reasoner: Advancing LLM Reasoning Across All Domains [NeurIPS25]
Official Repo for "TheoremExplainAgent: Towards Video-based Multimodal Explanations for LLM Theorem Understanding" [ACL 2025 oral]
Janus-Series: Unified Multimodal Understanding and Generation Models
Empowering RAG with a memory-based data interface for all-purpose applications!
Retrieval and Retrieval-augmented LLMs
An Open-Source Package for Information Retrieval
Official repo for "LongRAG: Enhancing Retrieval-Augmented Generation with Long-context LLMs".
See, hear, and interact on your local devices.
Official code for Paper "Mantis: Multi-Image Instruction Tuning" [TMLR 2024 Best Paper]