/ github.com / gke-labs / eval-runner

Comparisons

3 total
Comparison Verdict P(imp) Run deltas Tasks Baseline Run Candidate Run Completed Age
pr-395 UNCHANGED
UNCHANGED: all 2 task(s) unchanged
0.50
Quality 0.0 pp Tokens −64.1% Cost −$0.03 (−60.2%)
2/2 base-0a3e6b0 pr-395-4ea4529 2026-10-27 16:43 UTC 1h
pr-394 UNCHANGED
UNCHANGED: all 2 task(s) unchanged
0.50
Quality +12.5 pp Tokens +3.2% Cost +$0.0001 (+6.7%)
2/2 base-0a3e6b0 pr-394-7b89482 2026-10-27 16:38 UTC 1h
pr-391 INCONCLUSIVE
INCONCLUSIVE: 1 task(s) inconclusive; more samples recommended
0.63
Quality 0.0 pp Tokens −39.2% Cost −$0.02 (−36.3%)
2/2 base-0a3e6b0 pr-391-f3fda21 2026-10-27 15:48 UTC 2h