Seeing but Not Thinking: Routing Distraction in Multimodal Mixture-of-Experts
cs.CV, cs.AI, cs.CL
Submitted: 2026-04-09
Updated: 2026-11-15
Terminology
Sources
- Qwen3-VL Technical Report
- Multilingual Routing in Mixture-of-Experts
- xVerify: Efficient Answer Verifier for Reasoning Model Evaluations
- Training Verifiers to Solve Math Word Problems
- OmniSIFT: Modality-Asymmetric Token Compression for Efficient Omni-modal Large Language Models
- D$^2$HScore: Reasoning-Aware Hallucination Detection via Semantic Breadth and Depth Analysis in LLMs
- S3-CoT: Self-Sampled Succinct Reasoning Enables Efficient Chain-of-Thought LLMs
- Steering MoE LLMs via Expert (De)Activation
- Measuring Mathematical Problem Solving With the MATH Dataset
- Harder Tasks Need More Experts: Dynamic Routing in MoE Models
- The Platonic Representation Hypothesis
- SAFEx: Analyzing Vulnerabilities of MoE-Based LLMs via Stable Safety-critical Expert Identification
- Decoding Knowledge Attribution in Mixture-of-Experts: A Framework of Basic-Refinement Collaboration and Efficiency Analysis
- R2-T2: Re-Routing in Test-Time for Multimodal Mixture-of-Experts
- MoE-LLaVA: Mixture of Experts for Large Vision-Language Models
- SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization
- UI-S1: Advancing GUI Automation via Semi-online Reinforcement Learning
- Vision-Language Models Create Cross-Modal Task Representations
- Linearly Mapping from Image to Text Space
- Two Effects, One Trigger: On the Modality Gap, Object Bias, and Information Imbalance in Contrastive Vision-Language Models
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs