P^3IBench
Harbor: pendingA benchmark with 533 video-query samples evaluating proactive video assistants on perception, risk prediction, and intervention capabilities in real-life streaming scenarios.
- Samples
- 533
- Released
- 2026
- Scores
- 12
P-Bench
P-Bench P-Bench is a benchmark for evaluating statistical analysis agents on realistic, open-ended hypothesis-testing problems.
How to run
ltb · p-bench
$ npm i -g lettokenburn-cli$ ltb bench info p-bench
External harness. No Harbor adapter has been generated for this benchmark yet. `ltb bench info` prints the known links.
Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs
Scores
0 sourced entriesNo sourced scores recorded yet.
Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.
Reproduced by the community
All runsLoading community runs…
About
P-Bench
P-Bench is a benchmark for evaluating statistical analysis agents on realistic, open-ended hypothesis-testing problems. It contains 425 tasks spanning economics, biology, and medicine across 17 statistical test families. Given only a scientific hypothesis and a dataset, an agent must select an appropriate method, compute a p-value, and draw a conclusion.
Task composition of P-Bench. The inner ring shows the data domains, and the outer ring shows the…