Which Words a Model Fears: From a Refusal Direction to a Vocabulary-Wide Fear Lexicon
cs.CL
Submitted: 2026-11-02
Updated: 2026-11-02
Code: https://github.com/zboyr/which-words-a-model-fears
Terminology
Sources
- Automatic Pseudo-Harmful Prompt Generation for Evaluating False Refusals in Large Language Models
- Refusal in Language Models Is Mediated by a Single Direction
- Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models
- SCANS: Mitigating the Exaggerated Safety for LLMs via Safety-Conscious Activation Steering
- Training Verifiers to Solve Math Word Problems
- OR-Bench: An Over-Refusal Benchmark for Large Language Models
- Shortcut Learning in Deep Neural Networks
- Annotation Artifacts in Natural Language Inference Data
- Programming Refusal with Conditional Activation Steering
- GuardReasoner: Towards Reasoning-based LLM Safeguards
- Refusal in LLMs is an Affine Function
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
- Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference
- Safety Alignment Should Be Made More Than Just a Few Tokens Deep
- Steering Llama 2 via Contrastive Activation Addition
- XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models
- Navigating the OverKill in Large Language Models
- A StrongREJECT for Empty Jailbreaks
- Axiomatic Attribution for Deep Networks
- Universal Adversarial Triggers for Attacking and Analyzing NLP
Related papers
- ALARM: Audio-Language Alignment for Reasoning Models
- ThinkSum: Probabilistic reasoning over sets using large language models
- Coherence boosting: When your pretrained language model is not paying enough attention
- From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents
- Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization
- The Art of Asking: Multilingual Prompt Optimization for Synthetic Data