tml-epfl / llm-past-tense
Does Refusal Training in LLMs Generalize to the Past Tense? [NeurIPS 2024 Safe Generative AI Workshop (Oral)]
☆57Updated last month
Related projects ⓘ
Alternatives and complementary repositories for llm-past-tense
- [SafeGenAi @ NeurIPS 2024] Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates☆60Updated 3 weeks ago
- [NeurIPS 2024] Goldfish Loss: Mitigating Memorization in Generative LLMs☆74Updated this week
- Package to optimize Adversarial Attacks against (Large) Language Models with Varied Objectives☆64Updated 8 months ago
- Implementation of PaCE: Parsimonious Concept Engineering for Large Language Models (NeurIPS 2024)☆26Updated 2 weeks ago
- Improving Alignment and Robustness with Circuit Breakers☆154Updated last month
- Weak-to-Strong Jailbreaking on Large Language Models☆67Updated 9 months ago
- Code for the ICLR 2024 paper "How to catch an AI liar: Lie detection in black-box LLMs by asking unrelated questions"☆62Updated 5 months ago
- The repository contains code for Adaptive Data Optimization☆18Updated last month
- ☆49Updated last year
- Röttger et al. (2023): "XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models"☆63Updated 10 months ago
- Adversarial Attacks on GPT-4 via Simple Random Search [Dec 2023]☆42Updated 6 months ago
- ☆38Updated last year
- Replicating O1 inference-time scaling laws☆49Updated last month
- Codebase for Inference-Time Policy Adapters☆22Updated last year
- Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs☆39Updated 3 weeks ago
- Function Vectors in Large Language Models (ICLR 2024)☆119Updated last month
- [ICLR'24] RAIN: Your Language Models Can Align Themselves without Finetuning☆84Updated 5 months ago
- Is In-Context Learning Sufficient for Instruction Following in LLMs?☆25Updated 5 months ago
- WMDP is a LLM proxy benchmark for hazardous knowledge in bio, cyber, and chemical security. We also release code for RMU, an unlearning m…☆82Updated 6 months ago
- ☆78Updated last week
- Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".☆123Updated last month
- ☆49Updated 6 months ago
- [ICLR 2024] Provable Robust Watermarking for AI-Generated Text☆26Updated 11 months ago
- Safe Unlearning: A Surprisingly Effective and Generalizable Solution to Defend Against Jailbreak Attacks☆21Updated 4 months ago
- Official Repository for ACL 2024 Paper SafeDecoding: Defending against Jailbreak Attacks via Safety-Aware Decoding☆99Updated 4 months ago
- ☆31Updated last year
- [Arxiv 2024] Adversarial attacks on multimodal agents☆39Updated 4 months ago
- Privacy backdoors☆47Updated 6 months ago
- Landing Page for TOFU☆98Updated 5 months ago
- The official repository of the paper "On the Exploitability of Instruction Tuning".☆57Updated 9 months ago