← back to paper

Review history

arxiv: 2601.13981 · 2 revisions

VirtualCrime: Evaluating Criminal Potential of Large Language Models via Sandbox Simulation

1. 2026-05-21 CONDITIONAL LOW v0.9.0 novelty 6.0
   77383 ms 5762 in 1407 out 2026-05-21T15:17:01.790741+00:00
2. 2026-05-16 CONDITIONAL MODERATE v0.9.0 novelty 6.0
   26067 ms 5531 in 1206 out 2026-05-16T12:44:51.155106+00:00