List your company

Visual Spatial Reasoning on What's Up (Split A)

100AccuracyHuman Estimate
2242.2562.582.75Apr 29, 2025Jun 10, 2025Jul 22, 2025Sep 3, 2025Oct 15, 2025Nov 26, 2025Jan 8, 2026

Evaluation results

MethodLinks
Human Estimate2025.04100
SpaRE-7BModel size=7B2025.04100
GPT-4o2025.04100
Qwen2VL-7BModel size=7B2025.0499.5
LLaVA-NeXT-8BModel size=8B2025.0493.6
SpaRE-2BModel size=2B2025.0493.4
InternVL2-2BModel size=2B2025.0486.8
InternVL2-8BModel size=8B2025.0479.2
MiniCPM-V 4.5Prompting Strategy=General2026.0178.6
SpaceLLaVa2025.0475.5
GPT-4o-mini2025.0475
InternVL3.5 8BPrompting Strategy=Task-Aware2026.0164.8
InternVL3.5 8BPrompting Strategy=General2026.0163.3
XVLM 16M2026.0150.7
Qwen2VL-2BModel size=2B2025.0444.6
BLIP 14M2026.0138.8
MiniCPM-V 4.5Prompting Strategy=Task-Aware2026.0136.7
CLIP-ViT-B/322026.0131.3
Random2025.0425
Random2026.0125