thu-coai / SafetyBenchLinks
Official github repo for SafetyBench, a comprehensive benchmark to evaluate LLMs' safety. [ACL 2024]
☆223Updated 11 months ago
Alternatives and similar repositories for SafetyBench
Users that are interested in SafetyBench are comparing it to the libraries listed below
Sorting:
- ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors [EMNLP 2024 Findings]☆191Updated 8 months ago
- SC-Safety: 中文大模型多轮对抗安全基准☆135Updated last year
- S-Eval: Automatic and Adaptive Test Generation for Benchmarking Safety Evaluation of Large Language Models☆69Updated last month
- Flames is a highly adversarial benchmark in Chinese for LLM's harmlessness evaluation developed by Shanghai AI Lab and Fudan NLP Group.☆50Updated last year
- 【ACL 2024】 SALAD benchmark & MD-Judge☆147Updated 2 months ago
- [NAACL2024] Attacks, Defenses and Evaluations for LLM Conversation Safety: A Survey☆98Updated 9 months ago
- BeaverTails is a collection of datasets designed to facilitate research on safety alignment in large language models (LLMs).☆141Updated last year
- 面向中文大模型价值观的评估与对齐研究☆518Updated last year
- [ICLR 2024]Data for "Multilingual Jailbreak Challenges in Large Language Models"☆73Updated last year
- ☆86Updated 4 months ago
- R-Judge: Benchmarking Safety Risk Awareness for LLM Agents (EMNLP Findings 2024)☆76Updated 2 weeks ago
- [NAACL'24] Self-data filtering of LLM instruction-tuning data using a novel perplexity-based difficulty score, without using any other mo…☆368Updated 8 months ago
- Dataset and evaluation script for "Evaluating Hallucinations in Chinese Large Language Models"☆128Updated 11 months ago
- "他山之石、可以攻玉":复旦白泽智能发布面向国内开源和国外商用大模型的Demo数据集JADE-DB☆408Updated 2 months ago
- Do-Not-Answer: A Dataset for Evaluating Safeguards in LLMs☆250Updated 11 months ago
- 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。☆1,024Updated last year
- ☆45Updated last year
- ☆134Updated 9 months ago
- Chain of Attack: a Semantic-Driven Contextual Multi-Turn attacker for LLM☆31Updated 4 months ago
- Generative Judge for Evaluating Alignment☆238Updated last year
- Official github repo for AutoDetect, an automated weakness detection framework for LLMs.☆42Updated 11 months ago
- [arXiv:2311.03191] "DeepInception: Hypnotize Large Language Model to Be Jailbreaker"☆145Updated last year
- An LLM can Fool Itself: A Prompt-Based Adversarial Attack (ICLR 2024)☆86Updated 4 months ago
- JailBench:大型语言模型越狱攻击风险评测中文数据集 [PAKDD 2025]☆89Updated 3 months ago
- ☆32Updated 3 weeks ago
- The official implementation of our NAACL 2024 paper "A Wolf in Sheep’s Clothing: Generalized Nested Jailbreak Prompts can Fool Large Lang…☆109Updated 4 months ago
- 本文提出了一个基于“文心一言”的中国LLMs的安全评估基准,其中包括8种典型的安全场景和6种指令攻击类型。此外,本文还提出了安全评估的框架和过程,利用手动编写和收集开源数据的测试Prompts,以及人工干预结合利用LLM强大的评估能力作为“共同评估者”。☆25Updated last year
- [NeurIPS 2024 Oral] Aligner: Efficient Alignment by Learning to Correct☆176Updated 4 months ago
- InsTag: A Tool for Data Analysis in LLM Supervised Fine-tuning☆261Updated last year
- [NDSS'25 Best Technical Poster] A collection of automated evaluators for assessing jailbreak attempts.☆157Updated 2 months ago