Skip to content
← Back to explorer

Seeing but Not Thinking: Routing Distraction in Multimodal Mixture-of-Experts

Haolei Xu, Haiwen Hong, Hongxing Li, Rui Zhou, Yang Zhang, Longtao Huang, Hui Xue, Yongliang Shen, Weiming Lu, Yueting Zhuang · Apr 9, 2026 · Citations: 0

Expert Verification General
Open arXiv Find Implementation RSS feed Shortlist (0)

Data freshness

Extraction: Fresh

Check recency before relying on this page for active eval decisions. Use stale pages as context and verify against current hub results.

Metadata refreshed

Apr 9, 2026, 5:59 PM

Fresh

Extraction refreshed

Apr 10, 2026, 4:35 AM

Fresh

Extraction source

Persisted extraction

Confidence 0.45

Abstract

Multimodal Mixture-of-Experts (MoE) models have achieved remarkable performance on vision-language tasks. However, we identify a puzzling phenomenon termed Seeing but Not Thinking: models accurately perceive image content yet fail in subsequent reasoning, while correctly solving identical problems presented as pure text. Through systematic analysis, we first verify that cross-modal semantic sharing exists in MoE architectures, ruling out semantic alignment failure as the sole explanation. We then reveal that visual experts and domain experts exhibit layer-wise separation, with image inputs inducing significant routing divergence from text inputs in middle layers where domain experts concentrate. Based on these findings, we propose the Routing Distraction hypothesis: when processing visual inputs, the routing mechanism fails to adequately activate task-relevant reasoning experts. To validate this hypothesis, we design a routing-guided intervention method that enhances domain expert activation. Experiments on three multimodal MoE models across six benchmarks demonstrate consistent improvements, with gains of up to 3.17% on complex visual reasoning tasks. Our analysis further reveals that domain expert identification locates cognitive functions rather than sample-specific solutions, enabling effective transfer across tasks with different information structures.

Low-signal caution for protocol decisions

Use this page for context, then validate protocol choices against stronger HFEPX references before implementation decisions.

  • Extraction confidence is 0.45 (below strong-reference threshold).
  • No explicit evaluation mode was extracted from available metadata.
  • No benchmark/dataset or metric anchors were extracted.
Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub

HFEPX Relevance Assessment

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Best use

Background context only

Use if you need

Background context only.

Main weakness

Extraction confidence is 0.45 (below strong-reference threshold).

Trust level

Low

Eval-Fit Score

40/100 • Low

Treat as adjacent context, not a core eval-method reference.

Human Feedback Signal

Detected

Evaluation Signal

Weak / implicit signal

HFEPX Fit

Adjacent candidate

Extraction confidence: Low

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

Field Provenance & Confidence

Each key protocol field shows extraction state, confidence band, and data source so you can decide whether to trust it directly or validate from full text.

Human Feedback Types

partial

Expert Verification

Confidence: Low Source: Persisted extraction evidenced

Directly usable for protocol triage.

Evidence snippet: Multimodal Mixture-of-Experts (MoE) models have achieved remarkable performance on vision-language tasks.

Evaluation Modes

missing

None explicit

Confidence: Low Source: Persisted extraction missing

Validate eval design from full paper text.

Evidence snippet: Multimodal Mixture-of-Experts (MoE) models have achieved remarkable performance on vision-language tasks.

Quality Controls

missing

Not reported

Confidence: Low Source: Persisted extraction missing

No explicit QC controls found.

Evidence snippet: Multimodal Mixture-of-Experts (MoE) models have achieved remarkable performance on vision-language tasks.

Benchmarks / Datasets

missing

Not extracted

Confidence: Low Source: Persisted extraction missing

No benchmark anchors detected.

Evidence snippet: Multimodal Mixture-of-Experts (MoE) models have achieved remarkable performance on vision-language tasks.

Reported Metrics

missing

Not extracted

Confidence: Low Source: Persisted extraction missing

No metric anchors detected.

Evidence snippet: Multimodal Mixture-of-Experts (MoE) models have achieved remarkable performance on vision-language tasks.

Rater Population

partial

Domain Experts

Confidence: Low Source: Persisted extraction evidenced

Helpful for staffing comparability.

Evidence snippet: Multimodal Mixture-of-Experts (MoE) models have achieved remarkable performance on vision-language tasks.

Human Data Lens

  • Uses human feedback: Yes
  • Feedback types: Expert Verification
  • Rater population: Domain Experts
  • Unit of annotation: Unknown
  • Expertise required: General
  • Extraction source: Persisted extraction

Evaluation Lens

  • Evaluation modes:
  • Agentic eval: None
  • Quality controls: Not reported
  • Confidence: 0.45
  • Flags: ambiguous

Protocol And Measurement Signals

Benchmarks / Datasets

No benchmark or dataset names were extracted from the available abstract.

Reported Metrics

No metric terms were extracted from the available abstract.

Research Brief

Deterministic synthesis

Based on these findings, we propose the Routing Distraction hypothesis: when processing visual inputs, the routing mechanism fails to adequately activate task-relevant reasoning experts. HFEPX signals include Expert Verification with confidence 0.45. Updated from current HFEPX corpus.

Generated Apr 10, 2026, 4:35 AM · Grounded in abstract + metadata only

Key Takeaways

  • Based on these findings, we propose the Routing Distraction hypothesis: when processing visual inputs, the routing mechanism fails to adequately activate task-relevant reasoning…
  • Experiments on three multimodal MoE models across six benchmarks demonstrate consistent improvements, with gains of up to 3.17% on complex visual reasoning tasks.

Researcher Actions

  • Compare its human-feedback setup against pairwise and rubric hubs.
  • Identify benchmark choices from full text before operationalizing conclusions.
  • Verify metric definitions before comparing against your eval pipeline.

Caveats

  • Generated from title, abstract, and extracted metadata only; full-paper implementation details are not parsed.
  • Extraction confidence is probabilistic and should be validated for critical decisions.

Research Summary

Contribution Summary

  • Based on these findings, we propose the Routing Distraction hypothesis: when processing visual inputs, the routing mechanism fails to adequately activate task-relevant reasoning experts.
  • Experiments on three multimodal MoE models across six benchmarks demonstrate consistent improvements, with gains of up to 3.17% on complex visual reasoning tasks.

Why It Matters For Eval

  • Experiments on three multimodal MoE models across six benchmarks demonstrate consistent improvements, with gains of up to 3.17% on complex visual reasoning tasks.

Researcher Checklist

  • Pass: Human feedback protocol is explicit

    Detected: Expert Verification

  • Gap: Evaluation mode is explicit

    No clear evaluation mode extracted.

  • Gap: Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Gap: Benchmark or dataset anchors are present

    No benchmark/dataset anchor extracted from abstract.

  • Gap: Metric reporting is present

    No metric terms extracted.

Related Papers

Papers are ranked by protocol overlap, extraction signal alignment, and semantic proximity.

Need human evaluators for your AI research? Scale annotation with expert AI Trainers.

Post a Job Get a Quote