Benchmark RadarTSWAP Thai Wellness Benchmark
TSWAP Thai Wellness Benchmark is a retrieval-augmented question-answering benchmark for Thai traditional medicine and wellness advice. It includes 50 questions with gold document IDs, evaluated via Recall@5, and also releases production QA logs and a no-retrieval probe.
- Released
- 2026-08-24
- Readiness
- Inspectable
- Primary field
- Health & Life Sciences
Why it matters
The benchmark fills a gap in multilingual wellness and safety evaluation, providing a public dataset and standardized protocol to assess retrieval grounding and safety in LLM-based systems. It supports research in low-resource language NLP and responsible AI deployment.
Motivation
We present TSWAP, a deployed eight-language conversational wellness advisor grounded, via retrieval-augmented generation, in a verified knowledge base of Thai traditional medicine and certified wellness providers.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.