Benchmark track
Closed-Loop Learning
Task performance under stimulus-contingent feedback, versus matched no-feedback control.
Systems evaluated
0
Benchmark runs
0
Metrics
2
Tasks
2
Why this track exists
Closed-loop performance improvement (vs control) is the field's operational definition of stimulus-contingent adaptation at the task level. Not a claim about internal cognition.
Leaderboard
Only rows with run status published, provisional, or scored appear.
Metric definitions
| Metric | Description | Direction | Unit |
|---|---|---|---|
| Terminal performance delta | Final block performance in feedback condition minus no-feedback control. | higher_better | delta |
| Learning curve AUC | Area under the performance-over-time curve, feedback minus control. | higher_better | auc |
Scoring formula
Performance improvement over training (feedback minus no-feedback), z-scored against a batch-level null distribution.
Current methodology