centerforaisafety / wmdp
WMDP is a LLM proxy benchmark for hazardous knowledge in bio, cyber, and chemical security. We also release code for RMU, an unlearning method which reduces LLM performance on WMDP while retaining general capabilities.
☆105Updated 10 months ago
Alternatives and similar repositories for wmdp:
Users that are interested in wmdp are comparing it to the libraries listed below
- Röttger et al. (NAACL 2024): "XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models"☆91Updated 3 weeks ago
- Improving Alignment and Robustness with Circuit Breakers☆190Updated 5 months ago
- ☆53Updated 2 years ago
- ICLR2024 Paper. Showing properties of safety tuning and exaggerated safety.☆78Updated 10 months ago
- Official implementation of AdvPrompter https//arxiv.org/abs/2404.16873☆145Updated 10 months ago
- [ICLR 2025] Official Repository for "Tamper-Resistant Safeguards for Open-Weight LLMs"☆48Updated 3 weeks ago
- Finding trojans in aligned LLMs. Official repository for the competition hosted at SaTML 2024.☆109Updated 9 months ago
- LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces☆91Updated last year
- Weak-to-Strong Jailbreaking on Large Language Models☆72Updated last year
- A resource repository for representation engineering in large language models☆113Updated 4 months ago
- The official repository of the paper "On the Exploitability of Instruction Tuning".☆61Updated last year
- This is the starter kit for the Trojan Detection Challenge 2023 (LLM Edition), a NeurIPS 2023 competition.☆85Updated 10 months ago
- Improved Few-Shot Jailbreaking Can Circumvent Aligned Language Models and Their Defenses (NeurIPS 2024)☆58Updated 2 months ago
- Official implementation of ICLR'24 paper, "Curiosity-driven Red Teaming for Large Language Models" (https://openreview.net/pdf?id=4KqkizX…☆73Updated last year
- Official Repository for ACL 2024 Paper SafeDecoding: Defending against Jailbreak Attacks via Safety-Aware Decoding☆121Updated 8 months ago
- Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs☆66Updated 3 months ago
- AmpleGCG: Learning a Universal and Transferable Generator of Adversarial Attacks on Both Open and Closed LLM☆57Updated 4 months ago
- A lightweight library for large laguage model (LLM) jailbreaking defense.☆48Updated 5 months ago
- Adversarial Attacks on GPT-4 via Simple Random Search [Dec 2023]☆43Updated 10 months ago
- Package to optimize Adversarial Attacks against (Large) Language Models with Varied Objectives☆67Updated last year
- ☆30Updated 3 months ago
- Steering Llama 2 with Contrastive Activation Addition☆129Updated 10 months ago
- ☆38Updated last year
- [ICML 2024] Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications☆71Updated 3 weeks ago
- [ICLR'24] RAIN: Your Language Models Can Align Themselves without Finetuning☆90Updated 10 months ago
- ☆31Updated last year
- Benchmark evaluation code for "SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal" (ICLR 2025)☆48Updated 3 weeks ago
- A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicity.☆69Updated 2 weeks ago
- ☆42Updated last month
- A one-stop repository for large language model (LLM) unlearning. Supports TOFU, MUSE and is an easily extensible framework for new datase…☆183Updated last week