A free and strong UCI xiangqi engine

A platform for reproducible world model research and evaluation

[MLsys2026]: RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate, and 100% private RAG application on your personal device.

⚡FlashRAG: A Python Toolkit for Efficient RAG Research (WWW2025 Resource)

A fast, streaming-friendly BM25 search engine in Rust with mmap support

Emulation of the Drosophila Fly brain: Brian2, Brian2CUDA, PyTorch, NEST GPU, and neuromorphic chips

OpenResearcher: A Fully Open Pipeline for Long-Horizon Deep Research Trajectory Synthesis

📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG

Official JAX implementation of End-to-End Test-Time Training for Long Context

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models

A modular, primitive-first, python-first PyTorch library for Reinforcement Learning.

A general memory system for agents, powered by deep-research

Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.

The official repository of "A Comprehensive Survey on Reinforcement Learning-based Agentic Search: Foundations, Roles, Optimizations, Evaluations, and Applications".

Differentiable signal processing on the sphere for PyTorch

PyTorch-native post-training at scale

The best ChatGPT that $100 can buy.

LLM training in simple, raw C/CUDA

Storing long contexts in tiny caches with self-study

The open-source alternative to Claude Managed Agents

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.

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

🚀 Efficient implementations for emerging model architectures

DFloat11 [NeurIPS '25]: Lossless Compression of LLMs and DiTs for Efficient GPU Inference

Code to automatically prove or verify estimates in analysis

Everything about note management. All in Zotero.

NVIDIA cuML: GPU-Accelerated Machine Learning

Reference sparse coding implementations for efficient learning and inference.

Open-Sora: Democratizing Efficient Video Production for All

Tips for releasing research code in Machine Learning (with official NeurIPS 2020 recommendations)