List your company

Visual mathematical reasoning on MathVision mini

60.5AccuracyQwen3-VL-32B-Instruct
18.3829.31540.2551.185Mar 4, 2026Mar 22, 2026Apr 9, 2026Apr 28, 2026May 16, 2026Jun 3, 2026Jun 22, 2026

Evaluation results

MethodLinks
Qwen3-VL-32B-InstructContext Window=32K2026.0360.5
Qwen3-VL-32B-InstructContext Window=4K2026.0354.3
Qwen3-VL-8B-InstructContext Window=32K2026.0350
ReVisual-R1-7BTraining Stage=Baseline, Model Size=7B2026.0648.8
RL-Evol + Verifier (full VeriEvol)Training Stage=RL, Model Size=7B, Evolution Strategy=Evolved, Verifier Usage=Yes (HTV-Agent)2026.0647.37
Qwen3-VL-8B-InstructContext Window=4K2026.0345.7
RL-EvolTraining Stage=RL, Model Size=7B, Evolution Strategy=Evolved, Verifier Usage=No2026.0645.39
OpenMMReasoner-7BTraining Stage=Baseline, Model Size=7B2026.0643.6
RL-OriginTraining Stage=RL, Model Size=7B, Evolution Strategy=None, Verifier Usage=No2026.0641.45
VeriEvol-SFT (VeriEvol-SFT-Init)Training Stage=SFT, Model Size=7B, Evolution Strategy=Evolved, Verifier Usage=No2026.0639.8
OVR-7BTraining Stage=Baseline, Model Size=7B2026.0638.2
Phi-4-reasoning-vision-15B2026.0336.2
Seed-only SFTTraining Stage=SFT, Model Size=7B, Evolution Strategy=None, Verifier Usage=No2026.0636.16
Phi-4-reasoning-vision-15BProtocol=force nothink2026.0334.2
MM-Eureka-Qwen-7BTraining Stage=Baseline, Model Size=7B2026.0632.1
gemma-3-12b-it2026.0331.9
MMR1-Math-v0Training Stage=Baseline, Model Size=7B2026.0629
InternVL3-8BTraining Stage=Baseline, Model Size=8B2026.0628.6
VL-Rethinker-7BTraining Stage=Baseline, Model Size=7B2026.0628.4
Kimi-VL-A3B-Instruct2026.0328.3
WeThink-7BTraining Stage=Baseline, Model Size=7B2026.0627.2
OpenVLThinker-7BTraining Stage=Baseline, Model Size=7B2026.0626.9
VLAA-Thinker-7BTraining Stage=Baseline, Model Size=7B2026.0626.4
ThinkLite-VL-7BTraining Stage=Baseline, Model Size=7B2026.0624.6
Qwen2.5-VL-7B-InstructTraining Stage=Baseline, Model Size=7B2026.0621.8
Phi-4-mm-instruct2026.0320