PAST-Bench
Explore score history ↗PAST-Bench evaluates recursive self-improvement in personal AI agents by testing whether retained experience improves performance on future tasks. It spans 26 scenarios and 204 episodes across memory, procedural reuse, information gathering, and update capabilities, with matched persistence on/off controls.
- Released
- 2026-08-04
- Evaluates
- Agents, General AI, Language & Knowledge, Self-Evolution
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
PAST-Bench paper and code
No reported scores are on record for this benchmark yet.