← all runs
polyphony · gemini-2.5-flash
Mode
eval
Env
ci-frankenstein
Pass rate
40%
Result
fail
Duration
—
Cost
—
Tokens in/out
— / —
Git SHA
Started
2026-08-03 10:17
Cases (5, failures first)
eval:extractionscore 0.00fail
Output
F1=0.0 P=0.0 R=0.0 predicted=[]
eval:ingestionscore 0.00fail
Output
grounded=None/0 avg_chunks=None
eval:outlinescore 0.00fail
Output
nodes=0 structural_ok=False
⚖︎ Judge rationale
The response is an empty outline
eval:continuityscore 1.00pass
Output
detection_recall=1.0 fpr=1.0
eval:retrievalscore 0.85pass
Output
precision@k=0.8519 mrr=0.8704