apunkt/vestige
1
0
Fork 0
mirror of https://github.com/samvallad33/vestige.git synced 2026-07-26 23:51:02 +02:00
Code Issues Projects Releases Packages Wiki Activity
vestige/benchmarks/silent-rotation
History
Sam Valladares e751b427b4 benchmarks: add the sixth model (Kimi K2.7-code), including a trial Vestige lost 
I had set this run aside as too small. That was a self-serving exclusion: it is
the run where my own arm loses a trial on a model other than DeepSeek, and on
the same key (k_yildun) where it wins in the other run. Same model, same key,
opposite outcome. It belongs in the data.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-21 21:42:17 +07:00
..
harness benchmarks: add Silent Rotation, a fleet-convergence memory benchmark 2026-07-21 21:24:03 +07:00
results benchmarks: add the sixth model (Kimi K2.7-code), including a trial Vestige lost 2026-07-21 21:42:17 +07:00
tests benchmarks: add Silent Rotation, a fleet-convergence memory benchmark 2026-07-21 21:24:03 +07:00
EVIDENCE.md benchmarks: complete the Silent Rotation evidence set 2026-07-21 21:34:51 +07:00
FINDING.md benchmarks: add Silent Rotation, a fleet-convergence memory benchmark 2026-07-21 21:24:03 +07:00
QUOTES.md benchmarks: complete the Silent Rotation evidence set 2026-07-21 21:34:51 +07:00
SHOW-HN.md benchmarks: add Silent Rotation, a fleet-convergence memory benchmark 2026-07-21 21:24:03 +07:00