← back to paper

Review history

arxiv: 2607.28128 · 2 revisions

Rethinking LLM-Judged Helpfulness as a Pedagogy Signal: A Pre-Registered Audit Across Tutor Models

1. 2026-08-04 CONDITIONAL MODERATE v1.3.0-alltime-deepseek novelty 6.0
   146023 ms 30271 in 13660 out 2026-08-04T00:34:58.140754+00:00
2. 2026-07-31 CONDITIONAL HIGH v1.2.0-daily-grok45 novelty 6.5
   86962 ms 35575 in 4507 out 2026-07-31T16:58:54.014953+00:00