| Soft Prompt Threats: Attacking Safety Alignment and Unlearning in Open-Source LLMs through the Embedding Space | Feb 14, 2024 | Adversarial RobustnessSafety Alignment | CodeCode Available | 1 |
| Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications | Feb 7, 2024 | Safety Alignment | —Unverified | 0 |
| Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models | Feb 3, 2024 | Instruction FollowingSafety Alignment | CodeCode Available | 2 |
| MLLM-Protector: Ensuring MLLM's Safety without Hurting Performance | Jan 5, 2024 | Safety Alignment | CodeCode Available | 1 |
| Safety Alignment in NLP Tasks: Weakly Aligned Summarization as an In-Context Attack | Dec 12, 2023 | Question AnsweringSafety Alignment | CodeCode Available | 0 |
| RLHFPoison: Reward Poisoning Attack for Reinforcement Learning with Human Feedback in Large Language Models | Nov 16, 2023 | Backdoor AttackData Poisoning | —Unverified | 0 |
| Cognitive Overload: Jailbreaking Large Language Models with Overloaded Logical Thinking | Nov 16, 2023 | Safety Alignment | —Unverified | 0 |
| Trojan Activation Attack: Red-Teaming Large Language Models using Activation Steering for Safety-Alignment | Nov 15, 2023 | Red TeamingSafety Alignment | CodeCode Available | 1 |
| How Trustworthy are Open-Source LLMs? An Assessment under Malicious Demonstrations Shows their Vulnerabilities | Nov 15, 2023 | EthicsFairness | CodeCode Available | 0 |
| MART: Improving LLM Safety with Multi-round Automatic Red-Teaming | Nov 13, 2023 | Instruction FollowingRed Teaming | —Unverified | 0 |
