← back to paper

Review history

arxiv: 2605.13322 · 2 revisions

KamonBench: A Grammar-Based Dataset for Evaluating Compositional Factor Recovery in Vision-Language Models

1. 2026-05-20 UNVERDICTED LOW v0.9.0 novelty 7.0
   136243 ms 5728 in 1424 out 2026-05-20T21:35:18.290868+00:00
2. 2026-05-14 UNVERDICTED LOW v0.9.0 novelty 7.0
   26633 ms 5497 in 1171 out 2026-05-14T19:35:19.971470+00:00