List your company

Multimodal Understanding on MMStar (Pass@1)

60.3Pass@1 ScorePeBR-R1-7B
52.364854.424956.48558.5451Dec 13, 2025

Evaluation results

MethodLinks
PeBR-R1-7B2025.1260.3
PeRL-VLBackbone=Qwen2.5-VL-7B-Instruct2025.1259.95
SelfRewarded-R1-7B2025.1259.3
RL (Conditional Hard Gate)gamma=0, Backbone=Qwen2.5-VL-7B-Instruct2025.1258.56
RL (Verifiable Rewards)gamma=1, Backbone=Qwen2.5-VL-7B-Instruct2025.1258.12
R1ShareVL7B2025.1257.43
ThinkLite7B2025.1256.73
VL Cogito2025.1254.79
Base (Qwen2.5-VL-7B-Instruct)Backbone=Qwen2.5-VL-7B-Instruct2025.1252.67