List your company
ResearchFormerly WizWand

Multimodal Reasoning on MathVision (Metrics: FEI, AEI)

46.07FEIVRPRM-MiMo
23.720429.522735.32541.1273Aug 5, 2025
Updated 5mo ago

Evaluation results

MethodLinks
VRPRM-MiMo# Samples=53.6K, Explicit Reasoning (CoT)=true, RL Training=true2025.0846.0772.87
VRPRM-MiMo w/o CoT# Samples=53.6K, Explicit Reasoning (CoT)=false, RL Training=true2025.0843.469.99
VRPRM# Samples=53.6K, Explicit Reasoning (CoT)=true, RL Training=true2025.0842.9867.34
VRPRM-Qwen3# Samples=53.6K, Explicit Reasoning (CoT)=true, RL Training=true2025.0842.1370.48
VRPRM-Qwen3 w/o CoT# Samples=53.6K, Explicit Reasoning (CoT)=false, RL Training=true2025.0839.8970.15
VRPRM-MiMo w/o RL# Samples=3.6K, Explicit Reasoning (CoT)=true, RL Training=false2025.0839.3660.51
MiMo-VL-7B# Samples=unk2025.0838.2761.6
VRPRM-Qwen3 w/o RL# Samples=3.6K, Explicit Reasoning (CoT)=true, RL Training=false2025.0837.7857.69
Qwen3-VL-4B-Thinking# Samples=unk2025.0837.0861.64
Qwen2.5-VL-7B# Samples=unk2025.0836.9439.54
VRPRM-MiMo w/o RL & w/o CoT# Samples=3.6K, Explicit Reasoning (CoT)=false, RL Training=false2025.0835.8144.24
VRPRM-Qwen3 w/o RL & w/o CoT# Samples=3.6K, Explicit Reasoning (CoT)=false, RL Training=false2025.0834.9753.49
Qwen2.5-VL-72B# Samples=unk2025.0834.2741.88
VRPRM w/o RL# Samples=3.6K, Explicit Reasoning (CoT)=true, RL Training=false2025.0833.9961.82
VRPRM w/o RL & w/o CoT# Samples=3.6K, Explicit Reasoning (CoT)=false, RL Training=false2025.0833.1554.57
InternVL2.5-8B# Samples=unk2025.0830.6742.61
Gemini-2.0-Flash# Samples=unk2025.0830.4840.68
GPT-4o-mini# Samples=unk2025.0827.3935.1
VRPRM w/o CoT# Samples=53.6K, Explicit Reasoning (CoT)=false, RL Training=true2025.0826.8351.95
VisualPRM-8B# Samples=400K2025.0824.5862.91