← back to paper

Review history

arxiv: 2608.07419 · 2 revisions

Beyond Post-Hoc Temperature Scaling: Bilevel Optimization for LLM Calibration

1. 2026-08-15 CONDITIONAL MODERATE v1.4.0-alltime-deepseek-medium novelty 6.0
   134638 ms 28017 in 15680 out 2026-08-15T14:27:17.302008+00:00
2. 2026-08-10 CONDITIONAL MODERATE v1.4.0-daily-deepseek-medium novelty 6.0
   184926 ms 28018 in 22293 out 2026-08-10T04:55:55.287229+00:00