Papers
arxiv:2609.22153

SafeTune: A Unified Faithful Library for Auditing and Repairing Safety Drift in Fine-Tuned LLMs

Published on Aug 26
Authors:
,
,
,

Abstract

SafeTune provides a unified framework for comparing and applying four intervention paradigms to mitigate safety drift in fine-tuned large language models with shared evaluation and deployment support.

Methods for addressing safety drift in fine-tuned Large Language Models (LLMs) are scattered across incompatible implementations, lifecycle stages, and evaluation protocols, making them difficult to adopt and compare. We introduce SafeTune, a source-available library that unifies four intervention paradigms: post-hoc weight recovery, safety-constrained fine-tuning, gradient-based unlearning, and inference-time steering, alongside shared interpretability, evaluation, and deployment utilities. SafeTune provides a consistent configuration-driven workflow while preserving the distinct inputs and intervention points each paradigm requires. Its modular registry supports new methods, benchmarks, judges, models, and fine-tuning domains without redesigning the surrounding pipeline. We demonstrate SafeTune through controlled comparisons and finance and medical deployment case studies, showing how it characterizes safety drift, evaluates feasible interventions on common refusal-behavior and capability evaluations, and supports calibrated or layered mitigation.

Community

Sign up or log in to comment

Get this paper in your agent:

hf papers read 2609.22153
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2609.22153 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2609.22153 in a dataset README.md to link it from this page.

Spaces citing this paper 1

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.