List your company
ResearchFormerly WizWand

PT on Different-environment out-of-domain (Accuracy (%))

83.2Accuracy (OOD PT)Gemini 3 Flash
22.15238.00153.8569.699Jun 2, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Gemini 3 FlashModel Category=VQA Models, Evaluation Protocol=Zero-shot2026.0683.2
GPT-5Model Category=VQA Models, Evaluation Protocol=Zero-shot2026.0680.9
Gemini 2.5 FlashModel Category=VQA Models, Evaluation Protocol=Zero-shot2026.0671.4
Qwen3-VL-8BModel Category=VQA Models, Evaluation Protocol=Zero-shot2026.0664.1
GPT-5.2Model Category=VQA Models, Evaluation Protocol=Zero-shot2026.0663
Bagel + Mixed TrainingModel Category=Ours (fine-tuned BAGEL), Evaluation Protocol=Fine-tuned, Reasoning Strategy=IPT + Label-only mixture2026.0658.6
Bagel + IPTModel Category=Ours (fine-tuned BAGEL), Evaluation Protocol=Fine-tuned, Reasoning Strategy=Imaginative Perception Token2026.0657.5
Bagel (label-only)Model Category=Ours (fine-tuned BAGEL), Evaluation Protocol=Fine-tuned, Reasoning Strategy=Answer supervision only2026.0654.7
Bagel + Text CoTModel Category=Ours (fine-tuned BAGEL), Evaluation Protocol=Fine-tuned, Reasoning Strategy=Textual chain-of-thought2026.0652.2
InternVL3.5-8BModel Category=VQA Models, Evaluation Protocol=Zero-shot2026.0647.4
Qwen2.5-VL-7BModel Category=VQA Models, Evaluation Protocol=Zero-shot2026.0644.8
Bagel (base)Model Category=Ours (fine-tuned BAGEL), Evaluation Protocol=Zero-shot2026.0642.7
Janus-Pro-7BModel Category=Unified Models, Evaluation Protocol=Zero-shot2026.0635.3
Chameleon 7BModel Category=Unified Models, Evaluation Protocol=Zero-shot2026.0624.5