Research map / Vision-language and multimodal / Visual question answering
Benchmarking the Scientific Mind: A Pathology-Derived Biomedical VQA Benchmark for Complex Scientific Reasoning
Ziyu Zhao, Yiyang Liu, Yajiao Wang, Xiaotao Wang, Yang Li, Yuyang Peng, Jiaheng Zhou, Jinqiao Wang, Yingying Chen, Ge Yang, Haixin Wang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| score with free-form judge instead of LCR (removing LCR) for GPT-4o (ablation) | SORBE evaluation · SORBE | score: 60.0 -> 26.0 | Hurt |
| score with free-form judge instead of LCR for Gemini-2.5-Flash (ablation) | SORBE evaluation · SORBE | score: 68.3 -> 42.0 | Hurt |
| score with free-form judge instead of LCR for Fleming-VL-38B (ablation) | SORBE evaluation · SORBE | score: 45.8 -> 17.8 | Hurt |
| use LCR (harmonic mean of process and conclusion) instead of conclusion-only scoring for GPT-4o | SORBE evaluation · SORBE | score: 52.3 -> 26.0 | Hurt |
| domain-specific biomedical training (Fleming-VL-38B) versus generalist on SORBE | Fleming-VL-38B · SORBE | S_LCR: 89.8 Slake -> 17.8 SORBE | Hurt |
| increase size from Lingshu-7B to Lingshu-32B | Lingshu · SORBE | S_LCR: 19.2 -> 22.6 | Improved |
| increase size from Hulu-Med-7B to Hulu-Med-32B | Hulu-Med · SORBE | S_LCR: 18.2 -> 22.2 | Improved |
| use proprietary Gemini-2.5-Flash instead of best older open model Qwen2.5-VL-72B | MLLM choice · SORBE | S_LCR: 25.3 -> 42.0 | Improved |
Closest papers by meaning
- MicroVQA: A Multimodal Reasoning Benchmark for Microscopy-Based Scientific Research (CVPR 2025 · similarity 0.92)
- OpenMedReason: Scientific Reasoning Supervision for Medical Vision–Language Models (NeurIPS 2026 · similarity 0.91)
- Visual and Domain Knowledge for Professional-level Graph-of-Thought Medical Reasoning (ICML 2025 · similarity 0.91)
- Med-CMR: A Fine-Grained Benchmark Integrating Visual Evidence and Clinical Logic for Medical Complex Multimodal Reasoning (CVPR 2026 · similarity 0.90)
- Medical thinking with multiple images (ICLR 2026 · similarity 0.90)
- M3CoTBench: Benchmark Chain-of-Thought of MLLMs in Medical Image Understanding (ICLR 2026 · similarity 0.89)
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.
