mlsanityData to 14 Oct 2026

Visual Credit Audit for Multimodal Spatial Reasoning

arXiv 2607.27069 ↗First posted 29 Jul 2026cs.CV cs.AIcited by 1

On the map: LLM reasoning → Spatial and embodied reasoning

Abstract (arXiv)

Closed yes/no spatial benchmarks can reward a correct answer even when the image adds little support beyond no-image contexts. Under a fixed forced-choice interface, Visual Credit Audit (VCA) separates two estimands: whether the benchmark image gives the model's declared decision more support than text-only and blank controls, and whether the model responds to relation-specific visual evidence. The first audit is training- and label-free and does not require an answer flip. Applying labels yields dependence-credited correctness (D-CC); on correct items, it equals same-control gold-aligned positive gain, while prediction alignment extends the audit to errors. Across four open MLLMs and two spatial benchmarks, 12.73-26.25% of decisions are correct yet uncredited. Matched same-split image permutation reduces D-CC by 21.25-47.80 points, with every paired 95% interval above zero. Fixed-pixel relation contrasts and a 3x3 evidence-source factorial show why null controls cannot identify relation response. Among controlled correct-but-uncredited agreement decisions, response to relation reversal spans 81.57-100.00%, while 32.11% pooled change answer. Independently audited outcomes on 108 geometry-compatible edits provide a bounded natural-image correspondence check. VCA thereby decomposes benchmark success into correctness, additional image support, and relation-consistent response.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 18 of its 45 references that are in the map. First posted on arXiv 29 Jul 2026.

Builds on

18 of 45 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025Enhancing Spatial Reasoning through Visual and Textual Thinking28 Jul
  2. 2024GSR-BENCH: A Benchmark for Grounded Spatial Reasoning Evaluation via Multimodal LLMs19 Jun
  3. 2024SpatialPIN: Enhancing Spatial Reasoning Capabilities of Vision-Language Models through Prompting and Interacting 3D Priors18 Mar

Other references · 15

  1. 2025Qwen3-VL Technical Reportcited by 2,793
  2. 2024LLaVA-OneVision: Easy Visual Task Transfercited by 2,719
  3. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,224
  4. 2025InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiencycited by 1,510
  5. 2023Aligning Large Multimodal Models with Factually Augmented RLHFcited by 712
  6. 2023Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuningcited by 445
  7. 2026Ministral 3cited by 114
  8. 2025SpatialBench: Benchmarking Multimodal Large Language Models for Spatial Cognitioncited by 35
  9. 2024Look Before You Decide: Prompting Active Deduction of MLLMs for Assumptive Reasoningcited by 17
  10. 2026VisNec: Measuring and Leveraging Visual Necessity for Multimodal Instruction Tuningcited by 8
  11. 2026Do Vision-Language Models See or Guess? Measuring and Reducing Textual-Prior Reliance with a Phrasing-Controlled Benchmarkcited by 8
  12. 2026Do Images Speak Louder than Words? Investigating the Effect of Textual Misinformation in VLMscited by 7
3 more
  1. 2026Seeing without Looking: Do Vision-Language Benchmarks Really Test Vision?cited by 6
  2. 2026OmniMapBench: Benchmarking Visual-Centric Reasoning on Diverse Map Documentscited by 2
  3. 2026From Accuracy to Visual Dependence: Auditing and Filtering Modality Collapse in Traffic VideoQAcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Decodable but Misrouted: Sparse Features Uncover a Readout Gap in Vision-Language Models for Harmful Meme Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Trace, Verify, and Correct: A Training-Free Framework for Spatial Reasoning in Multimodal LLMs
  2. 2026SpatialChain: A Benchmark for Auditing Spatial Reasoning Faithfulness in VLMs
  3. 2026Frame Evidence:A Benchmark for Reference Frames and Visual Evidence in Multi-Image Spatial Reasoning
  4. 2026VisualFLIP: Do Predictions Depend on Task-Critical Visual Evidence in Multimodal Reasoning?
  5. 2026RECAP: Relation Evidence Calibration for Detecting Spatial Relation Hallucinations in Vision-Language Models
  6. 2026From Hallucination to Grounding: Diagnosing Visual Spatial Intelligence via CRISP
  7. 2026When More Is Less: A Systematic Analysis of Spatial and Commonsense Information for Visual Spatial Reasoning
  8. 2025VERIFY: A Benchmark of Visual Explanation and Reasoning for Investigating Multimodal Reasoning Fidelity