Skip to content

@thu-coai thu-coai

Change the repository type filter

All

    Repositories list

    92 repositories

    • TransferAttack

      Public
      Python
      •0•4•0•0•Updated Mar 5, 2025Mar 5, 2025

    • AISafetyLab

      Public
      AISafetyLab: A comprehensive framework covering safety attack, defense, evaluation and paper list.
      Python
      •
      MIT License
      •6•83•0•0•Updated Mar 3, 2025Mar 3, 2025

    • MAPS

      Public
      Official Implementation of ICLR25 paper "MAPS: Advancing Multi-modal Reasoning in Expert-level Physical Science"
      Python
      •0•1•0•0•Updated Mar 2, 2025Mar 2, 2025

    • LongSafety

      Public
      Python
      •
      MIT License
      •0•8•0•0•Updated Feb 26, 2025Feb 26, 2025

    • Agent-SafetyBench

      Public
      Python
      •0•16•0•0•Updated Feb 20, 2025Feb 20, 2025

    • ComplexBench

      Public
      Benchmarking Complex Instruction-Following with Multiple Constraints Composition (NeurIPS 2024 Datasets and Benchmarks Track)
      Python
      •
      MIT License
      •8•68•4•0•Updated Feb 20, 2025Feb 20, 2025

    • CharacterGLM-6B

      Public
      [EMNLP'24] CharacterGLM: Customizing Chinese Conversational AI Characters with Large Language Models
      Python
      •
      Apache License 2.0
      •36•451•2•0•Updated Jan 7, 2025Jan 7, 2025

    • SPaR

      Public
      Python
      •
      Apache License 2.0
      •3•42•1•0•Updated Dec 17, 2024Dec 17, 2024

    • CharacterBench

      Public
      [AAAI'25] CharacterBench: Benchmarking Character Customization of Large Language Models
      MIT License
      •0•13•1•0•Updated Dec 10, 2024Dec 10, 2024

    • MiniPLM

      Public
      [ICLR 2025] MiniPLM: Knowledge Distillation for Pre-Training Language Models
      Python
      •
      MIT License
      •6•34•3•0•Updated Nov 23, 2024Nov 23, 2024

    • MoralStory

      Public
      Python
      •0•17•1•0•Updated Nov 7, 2024Nov 7, 2024

    • OpenMEVA

      Public
      Benchmark for evaluating open-ended generation
      benchmarkevaluation-metricslanguage-generation
      Python
      •7•48•3•1•Updated Nov 6, 2024Nov 6, 2024

    • CodePlan

      Public
      0•6•1•0•Updated Oct 16, 2024Oct 16, 2024

    • ShieldLM

      Public
      ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors [EMNLP 2024 Findings]
      Python
      •
      MIT License
      •8•175•1•0•Updated Sep 29, 2024Sep 29, 2024

    • PICL

      Public
      Code for ACL2023 paper: Pre-Training to Learn in Context
      Python
      •
      MIT License
      •4•108•1•1•Updated Jul 26, 2024Jul 26, 2024

    • PsyQA

      Public
      一个中文心理健康支持问答数据集，提供了丰富的援助策略标注。可用于生成富有援助策略的长咨询文本。
      17•193•0•0•Updated Jul 21, 2024Jul 21, 2024

    • JailbreakDefense_GoalPriority

      Public
      [ACL 2024] Defending Large Language Models Against Jailbreaking Attacks Through Goal Prioritization
      Python
      •1•19•0•0•Updated Jul 9, 2024Jul 9, 2024

    • SafeUnlearning

      Public
      Safe Unlearning: A Surprisingly Effective and Generalizable Solution to Defend Against Jailbreak Attacks
      Python
      •1•24•2•0•Updated Jul 9, 2024Jul 9, 2024

    • CritiqueLLM

      Public
      Python
      •4•141•6•0•Updated Jul 1, 2024Jul 1, 2024

    • AutoDetect

      Public
      Official github repo for AutoDetect, an automated weakness detection framework for LLMs.
      Python
      •
      MIT License
      •1•41•0•0•Updated Jun 25, 2024Jun 25, 2024

    • BPO

      Public
      Python
      •
      Apache License 2.0
      •16•307•1•0•Updated Jun 24, 2024Jun 24, 2024

    • SafetyBench

      Public
      Official github repo for SafetyBench, a comprehensive benchmark to evaluate LLMs' safety. [ACL 2024]
      Python
      •
      MIT License
      •9•196•2•1•Updated Jun 24, 2024Jun 24, 2024

    • Emotional-Support-Conversation

      Public
      Data and codes for ACL 2021 paper: Towards Emotional Support Dialog Systems
      Python
      •
      Other
      •35•251•3•0•Updated Jun 19, 2024Jun 19, 2024

    • CrossWOZ

      Public
      A Large-Scale Chinese Cross-Domain Task-Oriented Dialogue Dataset
      Python
      •
      Apache License 2.0
      •115•673•3•1•Updated Jun 17, 2024Jun 17, 2024

    • ConvLab-2

      Public
      ConvLab-2: An Open-Source Toolkit for Building, Evaluating, and Diagnosing Dialogue Systems
      dialogue-systemstask-oriented-dialoguedialogue
      Python
      •
      Apache License 2.0
      •133•462•13•1•Updated Jun 17, 2024Jun 17, 2024

    • Safety-Prompts

      Public
      Chinese safety prompts for evaluating and improving the safety of LLMs. 中文安全prompts，用于评估和提升大模型的安全性。
      promptsafetyinstructionattack-defensechinese-languagellmprompt-engineeringchatgpt
      Apache License 2.0
      •84•942•1•0•Updated Feb 27, 2024Feb 27, 2024

    • Implicit-Toxicity

      Public
      Official Code for EMNLP 2023 paper: "Unveiling the Implicit Toxicity in Large Language Models""
      Python
      •0•10•0•0•Updated Nov 30, 2023Nov 30, 2023

    • Re3Dial

      Public
      Official Code for EMNLP 2023 paper: "Re3Dial: Retrieve, Reorganize and Rescale Conversations for Long-Turn Open-Domain Dialogue Pre-training"
      Python
      •0•6•1•0•Updated Oct 22, 2023Oct 22, 2023

    • DA-Transformer

      Public
      Official Implementation for the ICML2022 paper "Directed Acyclic Transformer for Non-Autoregressive Machine Translation"
      text-generationnatural-language-generationmachine-translationpytorch
      Python
      •
      Other
      •18•124•10•0•Updated Sep 10, 2023Sep 10, 2023

    • DAG-Search

      Public
      The beamsearch algorithm for DA-Transformer
      C++
      •
      Other
      •1•5•0•0•Updated Sep 10, 2023Sep 10, 2023