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

FlowVerse

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multimodal Mathematical ReasoningFlowVerse Text Centric
CoT Error38.1
38
Multi-modal Mathematical ReasoningFlowVerse raw
Performance70.4
20
Visual Mathematical ReasoningFlowVerse Vision Primary RP+EI+OQ
CoT-E Score69.4
20
Visual Mathematical ReasoningFlowVerse Vision Centric RP+EI+OQ
CoT Error32.6
20
Visual Mathematical ReasoningFlowVerse Vision Dense EI+OQ
CoT-E28.5
20
Visual Mathematical ReasoningFlowVerse Text Limited RP+EI+OQ
CoT Error35.7
20
Visual Mathematical ReasoningFlowVerse DI+RP+EI+OQ (All)
CoT Error34.4
20
Multimodal Mathematical ReasoningFlowVerse Vision Primary
CoT Error26.1
18
Multimodal Mathematical ReasoningFlowVerse Vision Centric
CoT Error39.6
18
Multimodal Mathematical ReasoningFlowVerse Vision Dense
CoT Error28.8
18
Multimodal Mathematical ReasoningFlowVerse Text Limited
CoT Error40.5
18
Multimodal Mathematical ReasoningFlowVerse All
CoT Error36.5
18
Mathematical Problem SolvingFlowVerse Text Plus DI+RP+EI+OQ
CoT Error41.7
17
Mathematical ReasoningFlowVerse Text Plus
CoT Error46.1
15
Visual Mathematical ReasoningFlowVerse (test)
CoT Error (All)37.8
13
Mathematical ReasoningFlowVerse
Accuracy70.4
10
Showing 16 of 16 rows