mlsanityData to 7 Oct 2026

PathGLS: Evaluating Pathology Vision-Language Models without Ground Truth through Multi-Dimensional Consistency

arXiv 2603.16113 ↗First posted 17 Mar 2026cs.CV cs.AIcited by 0

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

Abstract (arXiv)

Vision-Language Models (VLMs) offer significant potential in computational pathology by enabling interpretable image analysis, automated reporting, and scalable decision support. However, their widespread clinical adoption remains limited due to the absence of reliable, automated evaluation metrics capable of identifying subtle failures such as hallucinations. To address this gap, we propose PathGLS, a novel reference-free evaluation framework that assesses pathology VLMs across three dimensions: Grounding (fine-grained visual-text alignment), Logic (entailment graph consistency using Natural Language Inference), and Stability (output variance under adversarial visual-semantic perturbations). PathGLS supports both patch-level and whole-slide image (WSI)-level analysis, yielding a comprehensive trust score. Experiments on Quilt-1M, TCGA, REG2025, PathMMU and TCGA-Sarcoma datasets demonstrate the superiority of PathGLS. Specifically, on the Quilt-1M dataset, PathGLS reveals a steep sensitivity drop of 40.2% for hallucinated reports compared to only 2.1% for BERTScore. Moreover, validation against expert-defined clinical error hierarchies reveals that PathGLS achieves a strong Spearman's rank correlation of ρ=0.71 (p < 0.0001), significantly outperforming Large Language Model (LLM)-based approaches (Gemini 3.0 Pro: ρ=0.39, p < 0.0001). These results establish PathGLS as a robust reference-free metric. By directly quantifying hallucination rates and domain shift robustness, it serves as a reliable criterion for benchmarking VLMs on private clinical datasets and informing safe deployment. Code can be found at: https://github.com/My13ad/PathGLS

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 9 of its 13 references that are in the map. First posted on arXiv 17 Mar 2026.

Builds on

9 of 13 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024PathMMU: A Massive Multimodal Expert-Level Benchmark for Understanding and Reasoning in Pathology29 Jan
  2. 2023LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day1 Jun

Other references · 7

  1. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
  2. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,696
  3. 2021DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharingcited by 1,219
  4. 2018Object Hallucination in Image Captioningcited by 661
  5. 2025MedGemma Technical Reportcited by 398
  6. 2021RadGraph: Extracting Clinical Entities and Relations from Radiology Reportscited by 310
  7. 2023Quilt-1M: One Million Image-Text Pairs for Histopathologycited by 192

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Benchmarking Vision-Language Models for Automated Pathology Diagnosis and Report Generation
  2. 2026PathView-Bench: Can Multimodal Large Language Models Achieve Fine-grained Multiscale Understanding of Pathology Images?
  3. 2026Med-StepBench: A Hierarchical Reasoning Framework for Evaluating Hallucinations in Medical Vision-Language Models
  4. 2025PathFLIP: Fine-grained Language-Image Pretraining for Versatile Computational Pathology
  5. 2025PathoHR: Hierarchical Reasoning for Vision-Language Models in Pathology
  6. 2026Do Pathology Vision-Language Models Truly See Pathology?
  7. 2025Democratising Pathology Co-Pilots: An Open Pipeline and Dataset for Whole-Slide Vision-Language Modelling
  8. 2025MedHallTune: An Instruction-Tuning Benchmark for Mitigating Medical Hallucination in Vision-Language Models