Pith. sign in

← back to paper

Review history

arxiv: 2607.00572 · 2 revisions

HARC: Coupling Harmfulness and Refusal Directions for Robust Safety Alignment

  1. 2026-07-12 ACCEPT HIGH v1.1.0-grok45 novelty 7.0
    26812 ms 33168 in 2844 out 2026-07-12T09:25:48.734473+00:00
  2. 2026-07-02 UNVERDICTED LOW v0.9.1-grok novelty 6.0
    27511 ms 5763 in 1054 out 2026-07-02T12:55:17.583410+00:00