KW — F1-macro

F1-macro. The unweighted mean of per-class F1 scores, so every class counts equally regardless of frequency; PUMA’s primary triage metric (canonical baseline 0.5867).

Source: PUMA project documentation + raw benchmark results · Traceability: ANXAG-007,ANXAG-008,ANXAG-009 · Confidence: verified-at-primary-source