REVIEW 5 cited by
Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Safety aligned Large Language Models (LLMs) are vulnerable to harmful fine-tuning attacks -- a few harmful data mixed in the fine-tuning dataset can break the LLMs's safety alignment. While several defenses have been proposed, our evaluation shows that existing defenses fail \textit{when some specific training hyper-parameters are chosen} -- a large learning rate or a large number of training epochs in the fine-tuning stage can easily invalidate the defense. To this end, we propose Antidote, a post-fine-tuning stage solution, which remains \textbf{\textit{agnostic to the training hyper-parameters in the fine-tuning stage}}. Antidote relies on the philosophy that by removing the harmful parameters, the harmful model can be recovered from the harmful behaviors, regardless of how those harmful parameters are formed in the fine-tuning stage. With this philosophy, we introduce a one-shot pruning stage after harmful fine-tuning to remove the harmful weights that are responsible for the generation of harmful content. Despite its embarrassing simplicity, empirical results show that Antidote can reduce harmful score while maintaining accuracy on downstream tasks. Code is available at https://github.com/git-disl/Antidote.
Forward citations
Cited by 5 Pith papers
-
TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering
TamperBench systematically benchmarks 21 open-weight LLMs against nine fine-tuning and representation-space attacks and finds all of them can be tampered into producing harmful output while retaining capability.
-
LoX: Low-Rank Extrapolation Robustifies LLM Safety Against Fine-tuning
Low-rank extrapolation of an aligned model's weight update (LoX) reduces how much later fine-tuning erodes safety refusal behavior.
-
Vulnerability-Aware Alignment: Mitigating Uneven Forgetting in Harmful Fine-Tuning
Vulnerability-Aware Alignment splits safety training data into fragile and robust groups, then uses group robust optimization and adversarial perturbations, cutting harmful response rates after harmful fine-tuning by ...
-
Reshaping Representation Space to Balance the Safety and Over-rejection in Large Audio Language Models
A representation-space reshaping method improves LALM safety against harmful audio queries while keeping over-rejection low.
-
Guardians and Offenders: A Survey on Harmful Content Generation and Safety Mitigation of LLM
The submission's abstract promises an LLM safety survey, but the provided body is the opening page of an unrelated arithmetic-dynamics paper, so the artifact is internally inconsistent.
Discussion (0). Sign in to comment.