Experiment ·
HAIKUJUDGE
Does Claude Haiku 5.5 (effort low) agree with Opus judge.v1 at least as well as Sonnet-low on deep35-R450 and 4B-R450 so it can replace Sonnet-low as the L1 judge?
From the record
- What was tested
- HAIKUJUDGE
- Result
- Haiku ratio 0.566 vs Opus 0.634, Δ −0.069 (95 % CI −0.090, −0.049). Ranking deep35 − 4B: Haiku +0.007 [−0.018, +0.032] vs Opus +0.029 [−0.004, +0.063].
- Verdict
- adopted: shipped or recommended by its frozen rule
- Detail
- lever-adopted (judge protocol: Haiku 5.5-low adopted as L1 judge, 4/4 bars; addendum applied to L1-PROTOCOL.md:11)
- Set
- v0.1 EN
- Judge
- Claude Opus, Claude Haiku (L1), Claude Sonnet
- n
- 57 items x 2 legs (228 Haiku calls)
- Seeds
- judge seed 7; bootstrap seed 5
- Pre-registered
- yes, sha256-frozen
Linked facts · 1
- Haiku medium effort: second leg of HAIKUJUDGEHaiku medium effort: second leg of HAIKUJUDGEnot runKinda record, not a measurement (set, status, judge, n, seeds, date: not recorded)Caveatwhy closed: low effort already passed the agreement barSourcenot in the public repositories yetOpen the data file