← back to paper

Review history

arxiv: 2605.12925 · 2 revisions

AgentLens: Revealing The Lucky Pass Problem in SWE-Agent Evaluation

1. 2026-06-30 UNVERDICTED LOW v0.9.1-grok novelty 7.0
   36332 ms 5866 in 1385 out 2026-06-30T21:58:42.459891+00:00
2. 2026-05-14 CONDITIONAL LOW v0.9.0 novelty 7.0
   70456 ms 5641 in 1323 out 2026-05-14T18:46:32.801828+00:00