Steering Multimodal Large Language Models Decoding for Context-Aware Safety
Mirrored from arXiv — NLP / Computation & Language for archival readability. Support the source by reading on the original site.
Computer Science > Computation and Language
arXiv:2509.19212 (cs)
[Submitted on 23 Sep 2025 (v1), last revised 28 Aug 2026 (this version, v2)]
Title:Steering Multimodal Large Language Models Decoding for Context-Aware Safety
Authors:Zheyuan Liu, Zhangchen Xu, Guangyao Dou, Xiangchi Yuan, Zhaoxuan Tan, Radha Poovendran, Meng Jiang
View a PDF of the paper titled Steering Multimodal Large Language Models Decoding for Context-Aware Safety, by Zheyuan Liu and 6 other authors
View PDF
HTML (experimental)
Abstract:Multimodal Large Language Models (MLLMs) are increasingly deployed in real-world applications, yet their ability to make context-aware safety decisions remains limited. Existing methods often fail to balance oversensitivity (unjustified refusals of benign queries) and undersensitivity (missed detection of visually grounded risks), leaving a persistent gap in safety alignment. To address this issue, we introduce Safety-aware Contrastive Decoding (SafeCoDe), a lightweight and model-agnostic decoding framework that dynamically adjusts token generation based on multimodal context. SafeCoDe operates in two stages: (1) a contrastive decoding mechanism that highlights tokens sensitive to visual context by contrasting real and Gaussian-noised images, and (2) a global-aware token modulation strategy that integrates scene-level reasoning with token-level adjustment to adapt refusals according to the predicted safety verdict. Extensive experiments across diverse MLLM architectures and safety benchmarks, covering undersensitivity, oversensitivity, and general safety evaluations, show that SafeCoDe consistently improves context-sensitive refusal behaviors while preserving model helpfulness.
| Comments: | EMNLP 2026 Main |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2509.19212 [cs.CL] |
| (or arXiv:2509.19212v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2509.19212
arXiv-issued DOI via DataCite
|
Submission history
From: Zheyuan Liu [view email][v1] Tue, 23 Sep 2025 16:32:25 UTC (5,265 KB)
[v2] Fri, 28 Aug 2026 16:48:24 UTC (5,535 KB)
Full-text links:
Access Paper:
- View PDF
- HTML (experimental)
- TeX Source
View a PDF of the paper titled Steering Multimodal Large Language Models Decoding for Context-Aware Safety, by Zheyuan Liu and 6 other authors
References & Citations
Loading...
Bookmark
Bibliographic Tools
Code, Data, Media
Demos
Related Papers
About arXivLabs
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 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
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
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.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.