arXiv AI
SGM: Safety Glasses for Multimodal Large Language Models via Neuron-Level Detoxification
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv AI
Jun 30
Test-Time Detoxification without Training or Learning Anything
arXiv:2602. 02498v2 Announce Type: replace-cross Abstract: Large language models can produce toxic or inappropriate text even for benign inputs, creating risks when deployed at scale.
arXiv AI
Aug 11
Evolving Safety Landscape of Multi-modal Large Language Models: A Survey of Emerging Threats and Safeguards
arXiv:2608. 07535v1 Announce Type: cross Abstract: Multi-modal large language models (MLLMs) integrate heterogeneous modalities through modality alignment and fusion, enabling stronger understanding and reasoning.
arXiv Computation and Language
Sep 21
MME-Safety: A Fine-grained Benchmark for Safety Evaluation of MLLMs
arXiv:2609.20850v1 Announce Type: new Abstract: While Multimodal Large Language Models (MLLMs) show remarkable advancements, their cross-modal capabilities introduce complex vulnerabilities that easi...
arXiv Computation and Language
Aug 31
Steering Multimodal Large Language Models Decoding for Context-Aware Safety
The paper introduces Safety-aware Contrastive Decoding (SafeCoDe), a lightweight, model‑agnostic framework designed to improve context‑aware safety in Multimodal Large Language Models (MLLMs). SafeCoDe operates in two stages: a contrastive decoding step that highlights tokens sensitive to visual context by contrasting real and Gaussian‑noised images, and a global‑aware token modulation strategy that adjusts refusals based on scene‑level reasoning and predicted safety verdicts. Experiments across various MLLM architectures and safety benchmarks demonstrate that SafeCoDe consistently enhances context‑sensitive refusal behaviors while maintaining model helpfulness.
arXiv AI
Aug 10
CASA: Classification Augmented with Safety Attention for Robust Multimodal Alignment
arXiv:2604. 00310v2 Announce Type: replace-cross Abstract: Multimodal large-language models (MLLMs) often experience degraded safety alignment when harmful queries exploit cross-modal interactions.
arXiv Machine Learning
Aug 14
Adjustable Text-Guided Backdoor Attacks with Natural-Word Triggers on Multimodal Pretrained Models
arXiv:2604. 05809v2 Announce Type: replace-cross Abstract: This paper presents Text-Guided Backdoor (TGB), an adjustable backdoor attack against multimodal pretrained models that uses natural-word triggers, namely words that can naturally occur in ordinary textual inputs.