Evaluation Run 20260920T144826
Rubric 2026-09-20a · typesafe/jev-1.13 · 30 of 30 answered · 2.4s · $0.0024
No entry file was written. Results live in this run's ledger until Phase 4 adds the write path, after calibration sets the thresholds.
Per-dimension distribution
| Dimension | Type | Distribution | Mean confidence | Agrees |
|---|---|---|---|---|
answer_key_correct | noul | mean 0.93 · level 1 holds 30/30 FLAT | — | — |
difficulty | score | mean 1.35 · level 1 holds 16/30 | 0.58 | — |
distractor_pull | score | mean 1.51 · level 2 holds 19/30 | 0.54 | — |
engagement | score | mean 1.37 · level 2 holds 15/30 | 0.69 | — |
one_defensible_answer | noul | mean 0.80 · level 1 holds 22/23 FLAT | — | — |
passage_realism | score | mean 1.27 · level 1 holds 16/23 | 0.59 | — |
sat_choice_conventions | score | mean 1.83 · level 2 holds 29/30 FLAT | 0.76 | — |
sat_stem_conventions | score | mean 1.93 · level 2 holds 30/30 FLAT | 0.89 | — |
section | choice | reading 15/30 | 1.00 | 30/30 |
skill_code | choice | info-and-ideas 14/30 | 0.96 | — |
stem_self_contained | noul | mean 0.95 · level 1 holds 30/30 FLAT | — | — |
tested_construct_is_sat | noul | mean 0.96 · level 1 holds 30/30 FLAT | — | — |
A dimension marked FLAT holds more than 80% of the run on one level. That is either a clean bank or a dimension reporting a constant, and only a negative control tells them apart.
Run
| Served by | {'typesafe/jev-1.13-20260917 via TypeSafe': 30} |
| Input tokens | 58,016 |
| Failed | 0 |
| Preflight notes | 0 |
| Skipped entries | 0 |