List your company

Text-rich Visual Comprehension on SEED-Bench 2-Plus

70.22AccuracyReGuLaR-7B
46.476852.640958.80564.9691May 28, 2026

Evaluation results

MethodLinks
ReGuLaR-7BModel Category=Latent reasoning LVLMs2026.0570.22
LaserModel Category=Latent reasoning LVLMs2026.0570.05
DeepeyesModel Category=RL-based reasoning LVLMs2026.0569.08
Vision-R1Model Category=RL-based reasoning LVLMs2026.0568.95
MonetModel Category=Latent reasoning LVLMs2026.0565.88
Qwen2.5-VL-7BModel Category=General-purpose LVLMs2026.0565.31
LLaVA-OneVisionModel Category=General-purpose LVLMs2026.0561.22
PAPOModel Category=RL-based reasoning LVLMs2026.0554.11
LVRModel Category=Latent reasoning LVLMs2026.0547.39