List your company
ResearchFormerly WizWand

Real-world Visual Question Answering on MME-RealWorld-Lite (MMERW)

57AccuracyGPT-5-0807
29.096836.340943.58550.8291Oct 18, 2025Oct 21, 2025Oct 25, 2025Oct 29, 2025Nov 2, 2025Nov 6, 2025Nov 10, 2025
Updated 3mo ago

Evaluation results

MethodLinks
GPT-5-0807Model Category=Proprietary and Open-Source MLLMs2025.1157
GPT-4o-0513Model Category=Proprietary and Open-Source MLLMs2025.1151.6
SpatialThinker-30BTraining Dataset=STVQA-7K, Training Protocol=RL with Dense Rewards (Ours)2025.1149.2
SSL4RL-7B (Mask)Category=SSL4RL-7B, SSL Task=Mask2025.1049.03
SSL4RL-7B (Hard-Contrastive)Category=SSL4RL-7B, SSL Task=Hard-Contrastive2025.1048.61
SpatialThinker-7BTraining Dataset=STVQA-7K, Training Protocol=RL with Dense Rewards (Ours)2025.1148.3
Qwen2.5-VL-7B + SFTTraining Dataset=STVQA-7K, Training Protocol=SFT2025.1147.4
Claude-4-Sonnet-0514Model Category=Proprietary and Open-Source MLLMs2025.1146.9
Qwen2.5-VL-3B + Vanilla GRPOTraining Dataset=STVQA-7K, Training Protocol=Vanilla GRPO2025.1146.7
SpatialThinker-3BTraining Dataset=STVQA-7K, Training Protocol=RL with Dense Rewards (Ours)2025.1146.5
Qwen2.5-VL-7B + Vanilla GRPOTraining Dataset=STVQA-7K, Training Protocol=Vanilla GRPO2025.1146.3
Qwen3-VL-30BModel Category=Proprietary and Open-Source MLLMs2025.1145.8
Qwen2.5-VL-7BCategory=Base2025.1045.59
Claude-3.5-Sonnet-0620Model Category=Proprietary and Open-Source MLLMs2025.1145.2
Qwen2.5-VL 7B + NoisyGRPOModel=Qwen2.5-VL 7B, Training Strategy=+ NoisyGRPO2025.1044.6
VLAA-Thinker-7BModel Category=Proprietary and Open-Source MLLMs2025.1144.6
Qwen2.5-VL-7BModel Category=Proprietary and Open-Source MLLMs2025.1144.1
Qwen2.5-VL 3B + SFTModel=Qwen2.5-VL 3B, Training Strategy=+ SFT2025.1044
Qwen2.5-VL 3B + NoisyGRPOModel=Qwen2.5-VL 3B, Training Strategy=+ NoisyGRPO2025.1044
Qwen2.5-VL 7BModel=Qwen2.5-VL 7B, Training Strategy=Base2025.1043.6
Qwen2.5-VL 7B + SFTModel=Qwen2.5-VL 7B, Training Strategy=+ SFT2025.1043.5
Qwen2.5-VL-3B + SFTTraining Dataset=STVQA-7K, Training Protocol=SFT2025.1143
Qwen2.5-VL 7B + GRPOModel=Qwen2.5-VL 7B, Training Strategy=+ GRPO2025.1042.3
Qwen2.5-VL 3BModel=Qwen2.5-VL 3B, Training Strategy=Base2025.1042.1
Qwen2.5-VL-3BModel Category=Proprietary and Open-Source MLLMs2025.1141.9
Qwen2.5-VL 3B + GRPOModel=Qwen2.5-VL 3B, Training Strategy=+ GRPO2025.1040.8
PositionCategory=SSL4RL-3B2025.1038.19
JigsawCategory=SSL4RL-3B2025.1035.12
RotationCategory=SSL4RL-3B2025.1034.18
Qwen2.5-VL-3BCategory=Base2025.1032.41
ContrastiveCategory=SSL4RL-3B2025.1030.17