Model Context Protocol (MCP) server that lets AI assistants read Overleaf projects, parse LaTeX document structure, and push section-level edits back via Git. Compatible with Claude Desktop, Cursor…

Lightweight loop engineering state kernel for long-running AI agent teams. Agent-loop agnostic across Codex, Claude Code, and other coding agents, with durable goals, quota-aware auto-wake, executa…

Symphony turns project work into isolated, autonomous implementation runs, allowing teams to manage work instead of supervising coding agents.

A next.js web application that integrates AI capabilities with draw.io diagrams. This app allows you to create, modify, and enhance diagrams through natural language commands and AI-assisted visual…

The math library of Lean 4

Get the main content of any page as Markdown.

Uni-Agent is a framework for training long-horizon agents.

Unlimited OCR Works: Welcome the Era of One-shot Long-horizon Parsing.

code for "Tying the Loop - Tied Expert Layers in Mixture-of-Experts Language Models"

Code and Data for paper "GameCraft-Bench: Can Agents Build Playable Games End-to-End in a Real Game Engine?"

Lightweight coding agent that runs in your terminal

Skill to give Claude Code (and any coding agent) the ability to generate beautiful and practical Excalidraw diagrams.

DeepGEMM: clean and efficient BLAS kernel library on GPU

verl Zero-Mismatch Dense/MoE HuggingFace Rollout

A unified library of SOTA model optimization techniques like quantization, distillation, pruning, neural architecture search, speculative decoding, etc. It compresses deep learning models for downs…

A SOTA quantization algorithm for high-accuracy low-bit LLM inference, seamlessly optimized for CPU/XPU/CUDA, with multi-datatype support and full compatibility with vLLM, SGLang, and Transformers.

An asynchronous streaming data management module for efficient post-training.

Experimenting Heuristic Learning with ImageNet

MyPhoneBench: Do Phone-Use Agents Respect Your Privacy?

A MemAgent framework that can be extrapolated to 3.5M, along with a training framework for RL training of any agent workflow.

[ICLR'26] Stronger-MAS: A RL Framework for multi LLM agent system; [arxiv] MetaAgent-X: End-to-End Reinforcement Learning Automatic Multi-Agent Systems

Checkpoint-engine is a simple middleware to update model weights in LLM inference engines

Code and implementations for the paper "AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learning" by Zhiheng Xi et al.

Code for ICLR 2026 paper on "Understanding and Improving Shampoo and SOAP via Kullback-Leibler Minimization"

Implementation of GradLoc from the Tencent Hunyuan blog "Stabilizing RLVR via Token-level Gradient Diagnosis and Layerwise Clipping".