mlsanityData to 7 Oct 2026

Wasserstein Equilibrium Decoding for Reliable Medical Visual Question Answering

arXiv 2605.18313 ↗First posted 18 May 2026cs.CV cs.AIcited by 0

On the map: Multimodal LLMs / vision-language models → Medical vision-language models

Abstract (arXiv)

Small vision-language models (2-8B) are well-suited for clinical deployment due to privacy constraints, limited connectivity, and low-latency requirements favouring on-device or on-premise inference. However, their limited capacity exacerbates the generation of plausible but incorrect outputs. We extend game-theoretic decoding, previously restricted to text-only, closed-ended NLP tasks, to vision-language models for open-ended Medical VQA. We introduce a semantically aware Wasserstein stopping criterion that replaces lexical order matching, enabling convergence based on semantic consensus among near-synonymous candidate answers and avoiding unnecessary iterations caused by clinically equivalent ranking swaps. On VQA-RAD and PathVQA, we obtain consistent, statistically significant improvements over greedy and discriminative baselines. On VQA-RAD, we improve Qwen3-VL-2B by +3.5 percentage points (p < 0.01), surpassing the greedy 4B model, with similar trends at larger scales. On PathVQA, Gemma-3-4B with BDG matches MedGemma-4B under greedy decoding despite no domain-specific fine-tuning. At accuracy parity with classic BDG, the Wasserstein criterion reduces average convergence iterations by approximately 20%, improving inference efficiency while preserving the game-theoretic equilibrium behaviour. Code is available at https://github.com/luca-hagen/ Wasserstein-BDG-medical-VQA.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Medical vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 20 of its 37 references that are in the map. First posted on arXiv 18 May 2026.

Builds on

20 of 37 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024Vision-Language Models for Medical Report Generation and Visual Question Answering: A Review4 Mar
  2. 2023LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day1 Jun

Other references · 18

  1. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  2. 2022Survey of Hallucination in Natural Language Generationcited by 2,056
  3. 2024Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phonecited by 1,883
  4. 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,739
  5. 2024Gemma: Open Models Based on Gemini Research and Technologycited by 889
  6. 2023Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Modelscited by 695
  7. 2022Contrastive Decoding: Open-ended Text Generation as Optimizationcited by 598
  8. 2020PathVQA: 30000+ Questions for Medical Visual Question Answeringcited by 467
  9. 2023Med-Flamingo: a Multimodal Medical Few-shot Learnercited by 387
  10. 2020Self-Alignment Pretraining for Biomedical Entity Representationscited by 233
  11. 2024Rethinking the Bounds of LLM Reasoning: Are Multi-Agent Discussions the Key?cited by 153
  12. 2021Medical Visual Question Answering: A Surveycited by 93
6 more
  1. 2024Isambard-AI: a leadership class supercomputer optimised specifically for Artificial Intelligencecited by 59
  2. 2023The Consensus Game: Language Model Generation via Equilibrium Searchcited by 35
  3. 2025The Rise of Small Language Models in Healthcare: A Comprehensive Surveycited by 18
  4. 2024Memoryless Algorithmic Collusion: Sure to Fail, Slow to Fallcited by 3
  5. 2024Towards Edge General Intelligence via Large Language Models: Opportunities and Challengescited by 3
  6. 2024Algorithmic collusion in a two-sided market: A rideshare examplecited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026MedProb: Probing Internal Representations of Vision-Language Models for Medical Question Answering
  2. 2026MedLVR: Latent Visual Reasoning for Reliable Medical Visual Question Answering
  3. 2025Knowing or Guessing? Robust Medical Visual Question Answering via Joint Consistency and Contrastive Learning
  4. 2026Option-Aware Retrieval and Task-Specific VLM Adaptation for Medical VQA
  5. 2025DiN: Diffusion Model for Robust Medical VQA with Semantic Noisy Labels
  6. 2026LiteMedCoT-VL: Parameter-Efficient Adaptation for Medical Visual Question Answering
  7. 2026Do Medical Vision Language Models Actually See? A Counterfactual Grounding Framework and Hard-Negative Contrastive Training for Visually-Reliant Medical VLMs
  8. 2024Fusion of Domain-Adapted Vision and Language Models for Medical Visual Question Answering