THUSI-Lab / GameScaling

Scaling in Games: Continued Pre-Training for Embodied Agents in Diverse Virtual Worlds. Weights on Hugging Face.

lintsinghua / DeepAudit

DeepAudit：人人拥有的 AI 黑客战队，让漏洞挖掘触手可及。小白一键部署运行，自主协作审计 + 自动化沙箱 PoC 验证。支持 Ollama 私有部署 ，一键生成报告。支持中转站。​让安全不再昂贵，让审计不再复杂。

tsinghua-fib-lab / AgentSociety

AgentSociety 2 is a modern, LLM-native agent simulation platform designed for social science research and experimental design. It provides a flexible framework for creating and managing intelligent agents in simulated environments.

lintsinghua / claude-code-book

《御舆：解码 Agent Harness》42万字拆解 AI Agent 的Harness骨架与神经 —— Claude Code 架构深度剖析，15 章从对话循环到构建你自己的 Agent Harness。在线阅读网站：

bytedance / SALMONN

SALMONN family: A suite of advanced multi-modal LLMs

lintsinghua / paint-mv-skills

Agent skills that turn a song and its lyrics into a hand-painted watercolour music video (p5.brush + headless Chrome + ffmpeg), following PDoomVideo's pipeline

tuna / mirror-web

Source code of the web interface of https://mirrors.tuna.tsinghua.edu.cn/

TsinghuaC3I / Awesome-Memory-for-Agents

A Collection of Papers about Memory for Language Agents

BytedTsinghua-SIA / Open-MOPD

Open-MOPD: Diagnosing and Fixing Capability Imbalance in Multi-Teacher On-Policy Distillation

Tsinghua-MARS-Lab / SLAMFormer-Infinity

SLAMFormer-Infinity: Infinite SLAM Transformer for Unbounded Frontend and Backend Processing

siqiliu-tsinghua / tautology

Real analysis formalized in Lean 4 from nothing — the reals are constructed here. No mathlib, no external dependencies, no typeclasses.

tsinghua-fib-lab / AutoSOTA

—