AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

MedRepBench: Benchmarking Structured Understanding of Medical Report Images

Fangxin Shang, Yuan Xia, Dalu Yang, Yahui Wang, BingLinYang BingLinYang

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
apply GRPO with field-recall reward after SFTInternVL3-8B (Ours-SFT) · MedRepBench lab reports (no-OCR)average field recall (%): 73.31 -> 79.45Improved
apply GRPO with field-recall reward after SFTInternVL3-8B (Ours-SFT) · MedRepBench lab reports (OCR-assisted)average field recall (%): 82.37 -> 87.41Improved
apply GRPO after SFTInternVL3-8B · MedRepBench subjective (no-OCR)acceptability rate (%): 56.27 -> 60.64Improved
apply GRPO after SFTInternVL3-8B · MedRepBench subjective (no-OCR)excellence rate (%): 38.86 -> 42.45Improved
provide OCR text to the VLMQwen2.5-VL-32B · MedRepBench lab reportsRvalue recall (%): 76.32 -> 84.72Improved
provide OCR text to the VLMInternVL2.5-38B · MedRepBench lab reportsRabnormal recall (%): 66.01 -> 81.52Improved
provide OCR text to the VLMLLaMA-4-Maverick · MedRepBench lab reportsRabnormal recall (%): 75.83 -> 86.56Improved
provide OCR text to the VLMInternVL2.5-8B · MedRepBench lab reportsRname recall (%): 71.38 -> 85.64Improved
scale InternVL3 from 8B to 38BInternVL3 · MedRepBench subjective (no-OCR)acceptability rate (%): 48.24 -> 51.15No clear effect
scale Qwen2.5-VL from 7B to 32BQwen2.5-VL · MedRepBench subjective (no-OCR)excellence rate (%): 5.58 -> 51.40Improved
scale Qwen3 from 1.7B to 235B-A22B-2507Qwen3 (OCR+LLM) · MedRepBench lab reportsRabnormal recall (%): 69.04 -> 88.03Improved
scale InternVL3 from 8B to 38BInternVL3 · MedRepBench lab reports (no-OCR)Rvalue recall (%): 71.29 -> 64.83Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy