List your company
ResearchFormerly WizWand

High-Resolution Visual Reasoning on HRBench

0.7512AccuracyVision-R1
0.52760.585650.64370.70175Oct 2, 2025Oct 18, 2025Nov 4, 2025Nov 21, 2025Dec 8, 2025Dec 25, 2025Jan 11, 2026
Updated 8mo ago

Evaluation results

MethodLinks
Vision-R1Reasoning Paradigm=Tool-use & RL Enhanced Reasoning2026.010.7512
LaserReasoning Paradigm=Latent Reasoning2026.010.725
REWARDMAPModel=Qwen2.5-VL-7B-Instruct, Training Strategy=REWARDMAP2025.100.7125
Qwen2.5-VL-7B-Instruct + SFT -> RLModel=Qwen2.5-VL-7B-Instruct, Training Strategy=SFT -> RL Baseline2025.100.71
DeepEyesReasoning Paradigm=Tool-use & RL Enhanced Reasoning2026.010.6912
Qwen2.5-VL-7B-InstructModel=Qwen2.5-VL-7B-Instruct, Training Strategy=Base Model2025.100.6875
Qwen2.5-VL-7BReasoning Paradigm=Zero-Shot VLMs2026.010.6825
PAPOReasoning Paradigm=Tool-use & RL Enhanced Reasoning2026.010.6812
MonetReasoning Paradigm=Latent Reasoning2026.010.68
Qwen2.5-VL-3B-InstructModel=Qwen2.5-VL-3B-Instruct, Training Strategy=Reference2025.100.6625
VL-RethinkerReasoning Paradigm=Tool-use & RL Enhanced Reasoning2026.010.635
Qwen2.5-VL-32B-InstructModel=Qwen2.5-VL-32B-Instruct, Training Strategy=Reference2025.100.635
LLaVA-OneVisionReasoning Paradigm=Zero-Shot VLMs2026.010.63
InternVL3.5-8BReasoning Paradigm=Zero-Shot VLMs2026.010.5938
Kimi-VL-A3B-InstructModel=Kimi-VL-A3B-Instruct, Training Strategy=Reference2025.100.5538
LVRReasoning Paradigm=Latent Reasoning2026.010.5362