Our new X account is live! Follow @wizwand_team for updates

Metaphor Classification on MMSD Metaphor style 2.0 (test)

73.19Accuracy

Gemini 2.5 Flash

Updated 4d ago

Evaluation Results

Method	Links
Gemini 2.5 Flash 2026.01		73.19	0.76
SFT on Combined → GRPO (Combined) 2026.01		69.14	0.8
SFT on Combined → GRPO (Style specific) 2026.01		68.19	0.79
LLaMA 90B Vision Instruct 2026.01		67.56	0.76
Qwen2.5-VL 32B Instruct 2026.01		61.31	0.69
Gemma 3 27B 2026.01		58.44	0.64
Phi-4 Multimodal 2026.01		55.71	0.55
LLaVA 1.5 7B 2026.01		53.68	0.56