← back to paper

Review history

arxiv: 2606.27814 · 2 revisions

ATOD: Annealed Turn-Aware On-Policy Distillation for Multi-Turn Agentic Tasks

1. 2026-08-02 CONDITIONAL MODERATE v1.3.0-alltime-deepseek novelty 5.0
   156386 ms 16736 in 17264 out 2026-08-02T09:52:08.652995+00:00
2. 2026-06-29 UNVERDICTED LOW v0.9.1-grok novelty 6.0
   33200 ms 5772 in 1162 out 2026-06-29T04:27:33.051013+00:00