PsiDDx Clinical DDx: Differential Diagnosis Benchmark for LLMs

PsiDDx Clinical DDx is a benchmark dataset for evaluating large language models on clinical differential diagnosis tasks. It was published on Kaggle and features a model, Gemma-3-4B, reported to achieve a 75% win rate. The specific data volume, features, and creation details are not provided in the available metadata.

Use Cases

• Benchmarking LLM performance on clinical differential diagnosis (inferred from domain, verify after download)
• Training models for medical question-answering and reasoning (inferred from domain, verify after download)
• Evaluating the diagnostic accuracy of AI systems in simulated clinical scenarios (inferred from domain, verify after download)

Strengths

• Published on Kaggle, a major platform for data science.
• Includes a reported benchmark performance metric (75% win rate for Gemma-3-4B).

Limitations

• Metadata is minimal; actual content requires verification after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Kaggle