Pith. sign in

REVIEW 2 cited by

A safety realignment framework via subspace-oriented model fusion for large language models

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

arxiv 2405.09055 v1 pith:Q5FY5XLK submitted 2024-05-15 cs.CL

classification cs.CL
keywords safetymodelfine-tuningdownstreamfusionduringfine-tunedframework
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

The current safeguard mechanisms for large language models (LLMs) are indeed susceptible to jailbreak attacks, making them inherently fragile. Even the process of fine-tuning on apparently benign data for downstream tasks can jeopardize safety. One potential solution is to conduct safety fine-tuning subsequent to downstream fine-tuning. However, there's a risk of catastrophic forgetting during safety fine-tuning, where LLMs may regain safety measures but lose the task-specific knowledge acquired during downstream fine-tuning. In this paper, we introduce a safety realignment framework through subspace-oriented model fusion (SOMF), aiming to combine the safeguard capabilities of initially aligned model and the current fine-tuned model into a realigned model. Our approach begins by disentangling all task vectors from the weights of each fine-tuned model. We then identify safety-related regions within these vectors by subspace masking techniques. Finally, we explore the fusion of the initial safely aligned LLM with all task vectors based on the identified safety subspace. We validate that our safety realignment framework satisfies the safety requirements of a single fine-tuned model as well as multiple models during their fusion. Our findings confirm that SOMF preserves safety without notably compromising performance on downstream tasks, including instruction following in Chinese, English, and Hindi, as well as problem-solving capabilities in Code and Math.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Vulnerability-Aware Alignment: Mitigating Uneven Forgetting in Harmful Fine-Tuning

    cs.LG 2025-06 conditional novelty 6.0 of 10

    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 ...

  2. The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions

    cs.CL 2025-02 conditional novelty 6.0 of 10

    Safety refusal in Llama 3.1 8B is governed by a dominant activation direction plus smaller interpretable directions, and removing prompt tokens that activate these secondary directions can bypass fine-tuned safety.

Pith tools