Skip to content
AI News HubLIVE
Original source2 min read

SafeGene: Reusable Adapters for Transferable Safety Alignment

Summary

SafeGene introduces a reusable safety-adapter module that decouples safety capability from task-specific updates, enabling cross-task safety recovery within architecture-compatible model families. Experiments show reduced harmful response rates while maintaining downstream performance, outperforming representative safe adaptation methods.

SourcearXiv AIAuthor: Yanghan Wang, Zhiqiang Kou, Fu Feng, Jing Wang, Xin Geng
SafeGene: Reusable Adapters for Transferable Safety Alignment
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[2606.06519] SafeGene: Reusable Adapters for Transferable Safety Alignment

[Submitted on 2 Jun 2026]

Title:SafeGene: Reusable Adapters for Transferable Safety Alignment

View a PDF of the paper titled SafeGene: Reusable Adapters for Transferable Safety Alignment, by Yanghan Wang and 4 other authors

View PDF HTML (experimental)

Abstract:Open-weight LLMs are increasingly fine-tuned into customized assistants, but downstream fine-tuning can weaken safety alignment and make models more vulnerable to malicious prompts, even when the training data is not intentionally harmful. This creates a recurring safety recovery problem as target models are repeatedly updated with new task data or user interactions. We propose SafeGene, a reusable safety-adapter module designed for cross-task reuse within each architecture-compatible model family. Rather than treating safety recovery as a model-specific repair step, SafeGene treats safety capability as an independent, reusable adapter representation decoupled from task-specific updates. This representation is obtained from aligned--degraded model discrepancies, refined into task-transferable safety vectors through data-aware layer selection, and expressed in each downstream task-adapted model via few-shot layer-wise coefficient recalibration. Experiments across multiple model families, downstream tasks, and safety judges show that SafeGene-enhanced models reduce harmful response rates while maintaining downstream performance, outperforming representative safe adaptation methods in safety--utility trade-off.

Subjects:

Artificial Intelligence (cs.AI); Machine Learning (cs.LG)

Cite as: arXiv:2606.06519 [cs.AI]

(or arXiv:2606.06519v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2606.06519

arXiv-issued DOI via DataCite

Submission history

From: Yanghan Wang [view email] [v1] Tue, 2 Jun 2026 14:51:14 UTC (1,626 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled SafeGene: Reusable Adapters for Transferable Safety Alignment, by Yanghan Wang and 4 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.AI

new | recent | 2026-06

Change to browse by:

cs cs.LG

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • Models safety capability as an independent, reusable adapter representation rather than a model-specific repair step
  • Extracts safety vectors from aligned-degraded model discrepancies and refines them via data-aware layer selection
  • Applies the adapter through few-shot layer-wise coefficient recalibration in each downstream task-adapted model

Highlights and analysis are generated automatically and may contain errors. Check the original source.