List your company

Spatial Reasoning on VG-B SpatialScore-Hard

43.8AccuracyTVP (Generalize)
4.69614.8482535.152Dec 24, 2025

Evaluation results

MethodLinks
TVP (Generalize)Model Category=Visual Programming, Zero-shot Evaluation=true2025.1243.8
VADARModel Category=Visual Programming2025.1239.1
GPT-4oModel Category=Generic VLMs2025.1220.3
Molmo-7B-DModel Category=Generic VLMs2025.1212.5
LLaVA-OV-7B-ChatModel Category=Generic VLMs2025.129.4
Qwen2-VL-InstModel Category=Generic VLMs2025.127.8
SpaceMantisModel Category=Spatial-Finetuned VLMs2025.127.8
SpatialBot-3BModel Category=Spatial-Finetuned VLMs2025.126.2