yobro4619/vlm_ood_eval download history

yobro4619/vlm_ood_eval is a visual question answering dataset on the Hugging Face Hub. In the last 30 days it was downloaded 29 times (4 in the last 7 days), and 208 times in total. It ranks #391,467 among datasets by monthly downloads.

VLM OOD Eval (textual-prior reliance study) A 200-sample, open-ended, out-of-distribution evaluation set for measuring whether RL (GRPO) post-training of a vision-language model reduces textual-prior reliance (answering from text patterns instead of the image). It is a diverse OOD pro

Open yobro4619/vlm_ood_eval on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.