Share your thoughts, 1 month free Claude Pro on usSee more

Multimodal Mathematical Reasoning on FlowVerse Text Centric (Acc, CoT-E)

38.1CoT Error

Qwen-VL-Plus

Updated 1mo ago

Evaluation Results

Method	Links
Qwen-VL-Plus 2025.03		38.1	24.6
InfiMM-Math-7B 2025.03		43.8	38.1
InfiMM-Math-7B 2025.03		43.8	38.1
Gemini-Pro 2025.03		44.3	36.5
Math-LLaVA-13B 2025.03		45.1	39.3
MathLLaVA-13B 2025.03		45.1	39.3
InternVL2.5-8B 2025.03		49.2	41.3
InternVL2.5-8B 2025.03		49.2	41.3
Qwen-VL-Max 2025.03		49.8	42.1
Qwen-VL-Max 2025.03		49.8	42.1
MultiMath-7B 2025.03		50.6	44.8
MultiMath-7B 2025.03		50.6	44.8
SVE-Math-Qwen2.5-7B 2025.03		53.1	47.3
SVE-Math-Qwen2.5-7B 2025.03		53.1	47.3
GPT-4o-mini 2025.03		58.7	54.8
GPT-4o-mini 2025.03		58.7	54.8
VLM-R1-7B† 2025.03		59	54.2
VLM-R1-7B† 2025.03		59	54.2
Qwen2-VL-72B 2025.03		59.4	47.3
Qwen2-VL-72B 2025.03		59.4	47.3
Qwen2.5-VL-7B 2025.03		60.1	52.8
Qwen2.5-VL-7B 2025.03		60.1	52.8
Claude-3.5-Sonnet 2025.03		60.8	52.6
Claude-3.5-Sonnet 2025.03		60.8	52.6
GPT-4o 2025.03		61	56.8
GPT-4o 2025.03		61	56.8
MathFlow 2025.03		62	53.8
MathFlow 2025.03		62	53.8
InternVL2.5-78B 2025.03		66.1	62.7
InternVL2.5-78B 2025.03		66.1	62.7
Gemini-2.5-pro 2025.03		68.3	61.9
Gemini-2.5-pro 2025.03		68.3	61.9
GPT-4V 2025.03		69.1	57.1
GPT-4V 2025.03		69.1	57.1
GPT-5 2025.03		74.3	68.1
GPT-5 2025.03		74.3	68.1
MathFlow 2025.03		74.6	68.5
MathFlow 2025.03		74.6	68.5