LLM Safety Alignment

LLM: Large Language Model

Overview

Large language model safety alignment studies how to make models follow intended rules, avoid enabling harm, and remain useful when requests are ambiguous or adversarial. Common approaches include supervised safety fine-tuning and preference optimization such as reinforcement learning from human feedback (RLHF) and direct preference optimization (DPO), adversarial red-teaming, and inference-time controls such as refusal steering, prompt routing, or guard models. Evaluations use jailbreak and harmful-request suites such as HarmBench and JailbreakBench, alongside tests of over-refusal, multilingual and multimodal inputs, fine-tuning drift, and tool-using agents. A central challenge is achieving robust, context-sensitive safety without blocking benign help: refusals can fail under new formats, languages, or deployment pressures, while stronger safeguards can reduce utility and do not always predict real agent behavior.

Latest papers

CardsList

364 papers24 papers in the last four weeks, down 40% on the four weeks before. 0.2% of all new papers.

  1. Adaptive Helpfulness-Harmlessness Alignment with Preference Vectors

    Apr 27, 2025Ren-Wei Liang, Chin-Ting Hsu, Chan-Hung Yu +6LLM AlignmentLLM Safety Alignment

  2. Decoupled Alignment for Robust Plug-and-Play Adaptation

    Jun 3, 2024Haozheng Luo, Jiahao Yu, Wenxin Zhang +9LLM AlignmentLLM Safety Alignment

  3. Learning diverse attacks on large language models for robust red-teaming and safety tuning

    May 28, 2024Seanie Lee, Minsu Kim, Lynn Cherif +8Adversarial Attacks on LLMsLLM Red Teaming

  4. Eraser: Jailbreaking Defense in Large Language Models via Unlearning Harmful Knowledge

    Apr 8, 2024Weikai Lu, Ziqian Zeng, Jianwei Wang +5LLM Safety AlignmentLLM Jailbreak Attacks