Reproduce “NPCBench: A Clinical Apprenticeship Benchmark for Multimodal Care-Pathway Reasoning in Nasopharyngeal Carcinoma” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Pengkai Wang, Wei-Wei Zhang, Yan Li, Min Tang, Zhitian Hou, Zeyu Liu, Guanghao Zhu, Yuanyi Wang and 8 more
NPCBench tests multimodal models on nasopharyngeal carcinoma by linking pathology, MRI staging, and care decisions for each patient. It separates Care accuracy across individual decisions from Strict success, which requires correct pathology, MRI T stage, and every scored care answer. Across 21 models, Claude-4.6 led Strict at 23.0%; Gemini-3.1 led Care at 89.9%. CarePath scores linked tasks, not live clinical performance.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.