Pith. sign in

← back to paper

Review history

arxiv: 2508.13023 · 2 revisions

G$^2$RPO-A: Guided Group Relative Policy Optimization with Adaptive Guidance

  1. 2026-08-15 CONDITIONAL MODERATE v1.4.0-alltime-deepseek-medium novelty 5.0
    180397 ms 13849 in 19353 out 2026-08-15T17:17:02.110899+00:00
  2. 2026-08-05 REJECT MODERATE v1.4.0-alltime-deepseek-medium novelty 4.0
    133174 ms 13846 in 15025 out 2026-08-05T19:09:38.486915+00:00