作者位次 #4/4ACL 2026Poster· Natural Language Processing
M ed E inst: Benchmarking the Einstellung Effect in Medical LLM s through Counterfactual Differential Diagnosis
Wenting Chen · Guolin Huang · Wenxuan Wang · Zhongrui Zhu
摘要
Despite achieving high accuracy on medical benchmarks, LLMs exhibit the Einstellung Effect in clinical diagnosis—relying on statistical shortcuts rather than patient-specific evidence, causing misdiagnosis in atypical cases. Existing benchmarks fail to detect this critical failure mode. We introd...