Reading, Not Manipulating: Leveraging Router Logits for Multimodal Safety in MoE Vision-Language Models
cs.CL
Submitted: 2026-10-06
Updated: 2026-10-06
Sources
- Qwen3-VL Technical Report
- Learning to Detect Unseen Jailbreak Attacks in Large Vision-Language Models
- MSTS: A Multimodal Safety Test Suite for Vision-Language Models
- Linearly Decoding Refused Knowledge in Aligned Language Models
- Kimi-VL Technical Report
- DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding
- RouteHijack: Routing-Aware Attack on Mixture-of-Experts LLMs
- RASET: Router-Agnostic Safety-Critical Expert Tuning Exposes Localized Safety Enforcement Failures in Mixture-of-Experts LLMs
- InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering