List your company

Math Reasoning on MathVista GPS (test)

75.96AccuracyVeriGeo-beam5
20.569634.949849.3363.7102Jun 12, 2026

Evaluation results

MethodLinks
VeriGeo-beam5Params=7B, Data Scale=8.7k, Inference=beam52026.0675.96
VeriGeo-greedyParams=7B, Data Scale=8.7k, Inference=greedy2026.0675
GeoGen-SFT-7BParams=7B, Data Scale=224K2026.0674
MathCoder-VL-8BParams=8B, Data Scale=8.6M + 3M2026.0673.6
TR-CoT-8BParams=8B, Data Scale=87K2026.0673.1
InternVL2-76BParams=76B, Data Scale=-2026.0667.8
InternVL2.5-8BParams=8B, Data Scale=-2026.0667.8
MultiMath-7BParams=7B, Data Scale=300K2026.0666.8
MathCoder-VL-2BParams=2B, Data Scale=8.6M + 3M2026.0666.4
GPT-4oParams=Closed, Data Scale=-2026.0664.7
GeoGen-SFT-3BParams=3B, Data Scale=224K2026.0664.5
MathGLM-Vision-9BParams=9B, Data Scale=MathVL + VQA2026.0664.42
Claude-3.5-SonnetParams=Closed, Data Scale=-2026.0664.4
MAVIS-7BParams=7B, Data Scale=834K2026.0664.1
InternVL2-8BParams=8B, Data Scale=-2026.0662
Gemini-1.5-ProParams=Closed, Data Scale=-2026.0658.9
GPT-4-TurboParams=Closed, Data Scale=-2026.0658.3
Math-LLaVA-13BParams=13B, Data Scale=360K2026.0657.7
G-LLaVA-13BParams=13B, Data Scale=170K2026.0656.7
Gemini UltraParams=Closed, Data Scale=-2026.0656.3
InternVL2-26BParams=26B, Data Scale=-2026.0654.3
G-LLaVA-7BParams=7B, Data Scale=170K2026.0653.4
Qwen2.5-VL-7B-InstructParams=7B, Data Scale=-2026.0653.37
Claude-3-OpusParams=Closed, Data Scale=-2026.0652.9
GPT-4VParams=Closed, Data Scale=-2026.0650.5
Math-PUMA-Qwen2Params=7B/8B, Data Scale=-2026.0648.1
Qwen-VL-MaxParams=Closed, Data Scale=-2026.0646.1
Qwen2-VL-7B/8BParams=7B/8B, Data Scale=-2026.0640.9
Qwen-VL-PlusParams=Closed, Data Scale=-2026.0635.5
DeepSeek-VL-7BParams=7B, Data Scale=-2026.0628.4
LLaVA-1.5-13BParams=13B, Data Scale=-2026.0622.7