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

Compositional Generalization on Evaluation Dataset Unseen (Fold 1)

0.4818Score

DeepSeek V3

Updated 4d ago

Evaluation Results

Method	Links
DeepSeek V3 2026.01		0.4818
GPT-4o + FS* 2026.01		0.4668
Claude 3.5 Sonnet + FS* 2026.01		0.4627
LLaMA-3-8B 2026.01		0.4626
Gemma-7B 2026.01		0.451
GPT-4o 2026.01		0.4298
LLaMA-3-8B + COGLM 2026.01		0.4109
Qwen2.5-7B 2026.01		0.4079
Gemma-7B + COGLM 2026.01		0.3974
Claude 3.5 Sonnet 2026.01		0.3944
Mistral-7B 2026.01		0.388
Llama 3 (70B) + FS* 2026.01		0.3811
Mistral-7B + COGLM 2026.01		0.3786
Qwen2.5-7B + COGLM 2026.01		0.344
DeepSeek-7B 2026.01		0.3082
DeepSeek V3 + FS* 2026.01		0.3016
DeepSeek-7B + COGLM 2026.01		0.2953
Llama 3 (70B) 2026.01		0.2524