← back to paper

Review history

arxiv: 2606.31413 · 2 revisions

Learning to Select, Not Relearn: Hard-Routed Mixtures of Reasoning LoRAs

1. 2026-08-02 CONDITIONAL MODERATE v1.3.0-alltime-deepseek novelty 5.0
   252843 ms 21947 in 28294 out 2026-08-02T09:16:42.796196+00:00
2. 2026-07-01 UNVERDICTED LOW v0.9.1-grok novelty 6.0
   17903 ms 5785 in 1233 out 2026-07-01T06:02:30.057721+00:00