FBW-P07
Reward Association and Reversal Learning
Investigate how a model learns stimulus–reward associations and updates its choices when reward contingencies reverse.
Training preparationApplications open
Statistics temporarily unavailable
Project background
Separate pre-training, association, retention and reward reversal. Learning requires a recorded parameter or plastic-state change; changing the presentation alone is not learning. Begin with simple associations and short retention before extending the time scale.
This project connects a defined neural-computation question to a reproducible evaluation protocol and a community participation pathway. Results will be interpreted only within their recorded model, task and evidence scope.
Training goals
- Form a choice preference consistent with the reward rule.
- Compare learning efficiency and retention.
- Adapt when stimulus–reward associations are reversed.
Current foundation and milestones
Connectome and response computation are reusable. A learning mechanism and an association/reversal protocol must be selected and frozen before measured learning claims or curves can be published.
Next milestone: Select a learning mechanism and freeze association/reversal evaluation
- PlannedSelect a learning mechanism and freeze association/reversal evaluation
- PlannedImplement the versioned task and measurement interface
- PlannedEvaluate frozen conditions and independent reproduction
Content updated:
Technical approach
Use an explicitly documented reward-modulated plastic connection or constrained readout learning rule. Freeze the training/evaluation split and learning parameters. The current frozen integer model does not directly provide this capability: the learning mechanism is an independent version and must preserve historical scientific identities.
Method schematic
Task and evaluation design
Evaluation design
| Indicator | Definition | Direction | Units | Observation |
|---|---|---|---|---|
| Preference change | Compare matched choice preferences before and after training. | Compare matched conditions | preference difference | Awaiting evaluation |
| Learning trials | Trials required to meet a criterion frozen before training. | Lower is better | trials | Awaiting evaluation |
| Retention | Preference retained after the declared delay or interruption. | Higher is better | preference difference | Awaiting evaluation |
| Reversal adaptation | Trials needed to adapt after reward associations reverse. | Lower is better | trials | Awaiting evaluation |
No measured values have been published. Acceptance thresholds will be frozen in each task protocol before evaluation.
Expected deliverables
- Versioned learning configurations.
- Association and reversal process records.
- Independent evaluation and reproduction packages.
Ways to participate
- Learning-rule implementation
- Hyperparameter search
- Multiple-seed experiments
- Independent reproduction
Device requirements and task schedules will be matched by the community after runnable profiles have been evaluated. Unknown specifications are welcome in the application.
Apply to join training
Applications are received for community review and matching. Selection does not mean training has started.
References and related work
External sources provide technical background. They do not establish equivalent FBW capabilities.
View published research records ↗