THUKEG

AgentBench

A Comprehensive Benchmark to Evaluate LLMs as Agents (ICLR'24)

AgentTuning

AgentTuning: Enabling Generalized Agent Abilities for LLMs

CogDL

CogDL: A Comprehensive Library for Graph Deep Learning (WWW 2023)

GLM

GLM (General Language Model)

LongBench

LongBench v2 and LongBench (ACL 25'&24')

LongWriter

[ICLR 2025] LongWriter: Unleashing 10,000+ Word Generation from Long Context LLMs

P-tuning-v2

An optimized deep prompt tuning strategy comparable to fine-tuning across scales and tasks

slime

slime is a LLM post-training framework for RL Scaling.

T1

RL Scaling and Test-Time Scaling (ICML'25)

WebGLM

WebGLM: An Efficient Web-enhanced Question Answering System (KDD 2023)

THUDM/AgentBench

A Comprehensive Benchmark to Evaluate LLMs as Agents (ICLR'24)

THUDM/slime

slime is a LLM post-training framework for RL Scaling.

THUDM/LongWriter

[ICLR 2025] LongWriter: Unleashing 10,000+ Word Generation from Long Context LLMs

THUDM/WebGLM

WebGLM: An Efficient Web-enhanced Question Answering System (KDD 2023)

THUDM/SwissArmyTransformer

SwissArmyTransformer is a flexible and powerful library to develop your own Transformer variants.

THUDM/LongBench

LongBench v2 and LongBench (ACL 25'&24')

THUDM/ReST-MCTS

ReST-MCTS*: LLM Self-Training via Process Reward Guided Tree Search (NeurIPS 2024)

THUDM/LongCite

LongCite: Enabling LLMs to Generate Fine-grained Citations in Long-context QA

THUDM/WebRL

Building Open LLM Web Agents with Self-Evolving Online Curriculum RL

THUDM/LongAlign

[EMNLP 2024] LongAlign: A Recipe for Long Context Alignment of LLMs

THUDM/Android-Lab

THUDM/VisualAgentBench

Towards Large Multimodal Models as Visual Foundation Agents

THUDM/T1

RL Scaling and Test-Time Scaling (ICML'25)

THUDM/Awesome-Parameter-Efficient-Fine-Tuning-for-Foundation-Models

Parameter-Efficient Fine-Tuning for Foundation Models

THUDM/CogKit

Finetuning and inference tools for the CogView4 and CogVideoX model series.

THUDM/INFTY

INFTY Engine: An Optimization Toolkit to Support Continual AI

THUDM/TreeRL

TreeRL: LLM Reinforcement Learning with On-Policy Tree Search in ACL'25

THUDM/SWE-Dev

[ACL25' Findings] SWE-Dev is an SWE agent with a scalable test case construction pipeline.

THUDM/WhoIsWho

KDD'23 Web-Scale Academic Name Disambiguation: the WhoIsWho Benchmark, Leaderboard, and Toolkit

THUDM/DataSciBench

DataSciBench: An LLM Agent Benchmark for Data Science

THUDM/MoELoRA_Riemannian

Source code of paper: A Stronger Mixture of Low-Rank Experts for Fine-Tuning Foundation Models. (ICML 2025)

THUDM/DeepDive

DeepDive: Advancing Deep Search Agents with Knowledge Graphs and Multi-Turn RL

THUDM/scholar-profiling

THUDM/AndroidGen

THUDM/ReST-RL

Efficient Two-Stage Reinforcement Learning for LLMs

THUDM/BiPro

code and data for Paper: BIPro: Zero-shot Chinese Poem Generation via Block Inverse Prompting Constrained Generation Framework(ACL 2025 main)

THUDM/AlignMMBench

code, data and model for Paper: AlignMMBench: Evaluating Chinese Multimodal Alignment in Large Vision-Language Models (ACL'25 main)

THUDM/z-ai-sdk-typescript

Typescript SDK for Z.ai - Not yet released.

THUDM/BattleAgentBench

THUDM/MobileRL