MCP server for integrating OpenAI's Deep Research APIs and Hugging Face's Open Deep Research with Claude Code and other AI assistants

FlashInfer: Kernel Library for LLM Serving

RAGEN leverages reinforcement learning to train LLM reasoning agents in interactive, stochastic environments.

EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL

Training VLM agents with multi-turn reinforcement learning

verl: Volcano Engine Reinforcement Learning for LLMs

Resources and paper list for "Thinking with Images for LVLMs". This repository accompanies our survey on how LVLMs can leverage visual information for complex reasoning, planning, and generation.

GUI Grounding for Professional High-Resolution Computer Use

Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models know themselves through automated interpretability.

A research prototype of a human-centered web agent

Witness the aha moment of VLM with less than $3.

[CVPR 2025] Magma: A Foundation Model for Multimodal AI Agents

The official repo of the paper "MMLongBench Benchmarking Long-Context Vision-Language Models Effectively and Thoroughly"

Collect some World Models for Autonomous Driving (and Robotic) papers.

A framework for training world models with virtual environments, complete with annotated environment dataset (RetroAct), exploration agent (AutoExplore Agent), and GenieRedux-G - an implementation …

Building a comprehensive and handy list of papers for GUI agents

This repository introduce a comprehensive paper list, datasets, methods and tools for memory research.

Windows Agent Arena (WAA) 🪟 is a scalable OS platform for testing and benchmarking of multi-modal AI agents.

程序员在家做饭方法指南。Programmer's guide about how to cook at home (Simplified Chinese only).

open source interpretability platform 🧠

A Bionic Reading Extension for Zotero with Verbs and Nouns Highlight

[ACL 2024] Do Large Language Models Latently Perform Multi-Hop Reasoning?

Using Mechanistic Interpretability to Craft Adversarial Attacks against Large Language Models

The official repository for SkyLadder: Better and Faster Pretraining via Context Window Scheduling

Paper list for Efficient Reasoning.

Repository for the Q-Filters method (https://arxiv.org/pdf/2503.02812)

Jacobian SAEs for sparsifying LLM computation, rather than just representations

[ICLR 2025] Monet: Mixture of Monosemantic Experts for Transformers