List your company

Image-Text Reasoning on MathVerse

61.52AccuracyGRPO + ReMind
54.333656.199358.06559.9307Jun 2, 2026

Evaluation results

MethodLinks
GRPO + ReMindBase Model=Qwen3-VL-8B-Instruct2026.0661.52
DAPO + ReMindBase Model=Qwen3-VL-8B-Instruct2026.0660.37
DAPOBase Model=Qwen3-VL-8B-Instruct2026.0659.55
GRPOBase Model=Qwen3-VL-8B-Instruct2026.0659.52
RePOBase Model=Qwen3-VL-8B-Instruct2026.0659.39
RLEPBase Model=Qwen3-VL-8B-Instruct2026.0658.91
ExGRPOBase Model=Qwen3-VL-8B-Instruct2026.0656.89
Base ModelBase Model=Qwen3-VL-8B-Instruct2026.0654.61