Filling in the Clinical Gaps in Benchmark: Case for HealthBench for the Japanese medical system
It addresses the lack of Japanese medical benchmarks for LLM evaluation, though it is incremental as it adapts an existing benchmark rather than creating a new one.
This study investigated applying the HealthBench medical benchmark to Japan, finding that machine-translated versions caused performance drops in models like GPT-4.1 and failures in a Japanese-native model, with 5,000 scenarios revealing significant rubric mismatches requiring localization.
This study investigates the applicability of HealthBench, a large-scale, rubric-based medical benchmark, to the Japanese context. Although robust evaluation frameworks are essential for the safe development of medical LLMs, resources in Japanese are scarce and often consist of translated multiple-choice questions. Our research addresses this issue in two ways. First, we establish a performance baseline by applying a machine-translated version of HealthBench's 5,000 scenarios to evaluate two models: a high-performing multilingual model (GPT-4.1) and a Japanese-native open-source model (LLM-jp-3.1). Secondly, we use an LLM-as-a-Judge approach to systematically classify the benchmark's scenarios and rubric criteria. This allows us to identify 'contextual gaps' where the content is misaligned with Japan's clinical guidelines, healthcare systems or cultural norms. Our findings reveal a modest performance drop in GPT-4.1 due to rubric mismatches, as well as a significant failure in the Japanese-native model, which lacked the required clinical completeness. Furthermore, our classification shows that, despite most scenarios being applicable, a significant proportion of the rubric criteria require localisation. This work underscores the limitations of direct benchmark translation and highlights the urgent need for a context-aware, localised adaptation, a "J-HealthBench", to ensure the reliable and safe evaluation of medical LLMs in Japan.