Sweeperbench run database runs.csv is the append-only raw table. Each row describes one prediction, core evaluation, or preservation evaluation. schema.json documents the fields. Primary key: runid (UUID).
Dataset Card
Sweeperbench run database runs.csv is the append-only raw table. Each row describes one prediction, core evaluation, or preservation evaluation. schema.json documents the fields. Primary key: runid (UUID). Each top-level UUID directory is exactly that run, with matching metadata.json, artifact manifest, prompts, traces, verdicts and any collected recordings. Relationships: experimentid groups a launch; evaluations reference the exact predictionrunid. Model, harness
Excerpt from the card by Tony Chen.
Details
- Repository
- tonychenxyz/sweeperbench-traces
- Publisher
- Tony Chen
- Task category
- Not stated by the source
- Tags
- Not stated by the source
- Size category
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 59a0c3b8a4277a1798875d5bef93955a260c89eb
- Last updated
- 2026-09-25
License and Download
- License
- Not stated by the source
- Access
- No access gate
Released by Tony Chen through its official repository on Hugging Face.