Repository navigation

Model comparison: Haiku 4.5 vs Sonnet 5, run stability, close-call out of sample - #10

Merged

Lushenwar merged 4 commits into

mainfrom

feat/model-comparison

Oct 7, 2026

Merged

Model comparison: Haiku 4.5 vs Sonnet 5, run stability, close-call out of sample#10
Lushenwar merged 4 commits into
mainfrom
feat/model-comparison

Commits on Oct 7, 2026
- feat(core): optional SENTINEL_MODEL override (default claude-sonnet-5)
  
  Lushenwar
  and
  claude
  committed
  e74cce4
- feat(eval): meter token usage per call; label model comparisons by model
  
  Lushenwar
  and
  claude
  committed
  dfd2797
- feat(eval): Haiku 4.5 and Sonnet 5 re-run on eval set v2
  
  Lushenwar
  and
  claude
  committed
  3b2a4c7
- docs(metrics): model comparison, run stability, close-call rule out of sample
  
  Lushenwar
  and
  claude
  committed
  063efc5