List your company

Spatial Reasoning on VSR zero-shot (test)

63.67Accuracy (zero-shot)Llama-2 7B
50.98254.27657.5760.864Dec 22, 2025

Evaluation results

MethodLinks
Llama-2 7BLLM backbone=Llama-2, Model Size=7B2025.1263.67
Llama-2 Chat 7BLLM backbone=Llama-2, Model Size=7B, Variant=Chat2025.1261.8
Moxin-7BLLM backbone=Moxin-7B, Model Size=7B2025.1260.8
Mistral v0.1 7BLLM backbone=Mistral, Model Size=7B, Variant=v0.12025.1258.5
Mistral Instruct v0.1 7BLLM backbone=Mistral, Model Size=7B, Variant=Instruct v0.12025.1257.8
LLaVa v1.5 7B (Base)LLM backbone=LLaVa v1.5, Model Size=7B, Variant=Base2025.1251.47