LayerRAG-Bench
Explore score history ↗Evaluates cross-layer reliability of agentic RAG systems on 240 tasks across 8 enterprise domains, with 9 fault scenarios and 2 contract modes; measures success at evidence, tool-contract, authorization, and session-state layers.
- Released
- 2026-07-29
- Evaluates
- General AI, Information retrieval, Language & Knowledge, cs.CL
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
LayerRAG-Bench paper and code
No reported scores are on record for this benchmark yet.