Open-source platform to build and deploy AI agent workflows.

12 weeks, 26 lessons, 52 quizzes, classic Machine Learning for all

📚 《从零开始构建智能体》——从零开始的智能体原理与实践教程

An AI-powered Texas Hold'em Poker framework driven by Large Language Models

[ICLR 2025] 🧬 RegMix: Data Mixture as Regression for Language Model Pre-training (Spotlight)

[EMNLP 2025 Main] Agent-as-Judge for Factual Summarization of Long Narratives

verl: Volcano Engine Reinforcement Learning for LLMs

微舆：人人可用的多Agent舆情分析助手，打破信息茧房，还原舆情原貌，预测未来走向，辅助决策！从0实现，不依赖任何框架。

An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & TIS & vLLM & Ray & Async RL)

[CVPR2024 Highlight] VBench - We Evaluate Video Generation

Whether you're compiling kernels, training models, or just waiting for sleep 600 to finish, JobDone makes sure you never miss the moment your job ends—successfully, tragically, or somewhere in betw…

The best ChatGPT that $100 can buy.

12 Weeks, 24 Lessons, AI for All!

An easy-to-use Python framework for testing the robustness of models as evaluators

[ACL 2025] Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atoms

😼 优雅地使用基于 clash/mihomo 的代理环境

This repository collects all relevant resources about interpretability in LLMs

An easy-to-use Python framework to generate adversarial jailbreak prompts.

Landing page for MIB: A Mechanistic Interpretability Benchmark

这是一个简单的技术科普教程项目，主要聚焦于解释一些有趣的，前沿的技术概念和原理。每篇文章都力求在 5 分钟内阅读完成。

Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".

Text-audio foundation model from Boson AI

Performant framework for training, analyzing and visualizing Sparse Autoencoders (SAEs) and their frontier variants.

Code for USENIX Security 2025 paper "JBShield: Defending Large Language Models from Jailbreak Attacks through Activated Concept Analysis and Manipulation"

code for EMNLP 2024 paper: Neuron-Level Knowledge Attribution in Large Language Models