← back to paper

Review history

arxiv: 2605.22297 · 2 revisions

One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs

1. 2026-06-30 UNVERDICTED LOW v0.9.1-grok novelty 5.0
   39778 ms 5837 in 1376 out 2026-06-30T17:28:45.970677+00:00
2. 2026-05-22 CONDITIONAL MODERATE v0.9.0 novelty 6.0
   24812 ms 5788 in 1410 out 2026-05-22T07:40:59.655959+00:00