List your company

Meme Generation on GOAT-G (test)

60.1BERTScore F1Gemini-3-Flash
37.01243.0064954.994Jun 2, 2026

Evaluation results

MethodLinks
Gemini-3-Flashzero-shot=true2026.0660.143.7
Claude-Sonnet-4.6zero-shot=true2026.065841.2
Intent ProjectionBackbone=InternVL3-8B, Strategy=full2026.0653.133.8
GPT-5.4-minizero-shot=true2026.0652.433.8
Intent ProjectionBackbone=Qwen3-VL-8B, Strategy=full2026.065232.3
InternVL3-8BStrategy=+ SFT + Chain + GRPOflat2026.0651.231.9
Qwen3-VL-8BStrategy=+ SFT + Chain + GRPOflat2026.0650.230.4
InternVL3-8BStrategy=+ SFT + Chain2026.0649.630.3
Qwen3-VL-8BStrategy=+ SFT + Chain2026.0648.628.9
InternVL3-8BStrategy=+ SFT2026.0648.329.2
Intent ProjectionBackbone=LLaVA-OneVision-7B, Strategy=full2026.0647.428.3
Qwen3-VL-8BStrategy=+ SFT2026.0647.427.8
LLaVA-OneVision-7BStrategy=+ SFT + Chain + GRPOflat2026.0645.626.6
InternVL3-8BStrategy=+ CoT2026.0645.526
Qwen3-VL-8BStrategy=+ CoT2026.064424.6
LLaVA-OneVision-7BStrategy=+ SFT + Chain2026.0643.925.1
InternVL3-8BStrategy=Base2026.0643.924.1
LLaVA-OneVision-7BStrategy=+ SFT2026.0642.624
Qwen3-VL-8BStrategy=Base2026.0642.122.5
LLaVA-OneVision-7BStrategy=+ CoT2026.0639.420.8
LLaVA-OneVision-7BStrategy=Base2026.0637.919.1