RES — A three-seed validation {42,123,456} of the triage baseline produced…
A three-seed validation {42,123,456} of the triage baseline produced identical F1-macro = 0.5831 (0.0000 std-dev, 0% CV); run-to-run variability is confined to runtime metrics (~2% duration, ~4% energy), not task metrics.
Source: puma docs/results/multi_seed_baseline.md · Traceability: corpus unit
PUMA-020· Confidence: repo-backed