Pith. sign in

← back to paper

Review history

arxiv: 2605.21602 · 2 revisions

Benchmarking and Improving Monitors for Out-Of-Distribution Alignment Failure in LLMs

  1. 2026-06-30 CONDITIONAL MODERATE v0.9.1-grok novelty 6.0
    42666 ms 5817 in 1326 out 2026-06-30T17:14:43.537121+00:00
  2. 2026-05-22 UNVERDICTED LOW v0.9.0 novelty 6.0
    29907 ms 5778 in 1164 out 2026-05-22T09:34:20.324048+00:00