DD-011 v3
Threshold and dynamic experiment extension
Eight synthetic treatments separate threshold portfolios, history visibility, and fixed-versus-stopping action horizons. Six appended hypotheses are model-derived design inputs, not behavioral findings.
Scroll horizontally to see every column.
| Cell | Threshold | Portfolio rule | History | Horizon |
|---|---|---|---|---|
| T29-threshold-paired-planner | two | paired_planner | not_dynamic | one_shot |
| T30-threshold-common-mode | two | common_mode | not_dynamic | one_shot |
| T31-threshold-private-following | two | private_following | not_dynamic | one_shot |
| T32-threshold-team-tokens | two | team_tokens | not_dynamic | one_shot |
| T33-dynamic-bayes-hidden-fixed | none | none | hidden | fixed_budget |
| T34-dynamic-bayes-visible-fixed | none | none | visible | fixed_budget |
| T35-dynamic-planner-hidden-fixed | none | dynamic_planner | hidden | fixed_budget |
| T36-dynamic-planner-hidden-stopping | none | dynamic_planner | hidden | stopping |
Scroll horizontally to see every column.
| ID | Question | Outcome | Multiplicity family |
|---|---|---|---|
| H15 | Does paired threshold allocation improve discovery over common-mode crowding? | discovery | threshold |
| H16 | Do team-token recommendations improve threshold portfolio efficiency over common mode? | threshold_portfolio_efficiency | threshold |
| H17 | Does private following increase viable-team count over common mode? | viable_team_count | threshold |
| H18 | Does hidden history preserve discovery relative to visible action history? | discovery | dynamic |
| H19 | Does the registered stopping policy save actions relative to a fixed planner horizon? | action_savings | dynamic |
| H20 | Does hidden-history planner allocation improve discovery over hidden-history Bayes behavior? | discovery | dynamic |
Download v3 design · Download all retained calibration failures