This PR enhances the dataset card for CogIP-Bench by:
- Adding explicit links to the official paper ([From Pixels to Feelings: Aligning MLLMs with Human Cognitive Perception of Images](https://huggingface.co/papers/2511.22805)), the project page ([https://follen-cry.github.io/MLLM-Cognition-project-page/](https://follen-cry.github.io/MLLM-Cognition-project-page/)), and the GitHub repository ([https://github.com/Follen-cry/MLLM_Cognition_Alignment](https://github.com/Follen-cry/MLLM_Cognition_Alignment)).
- Updating the `task_categories` in the metadata from `question-answering` to `image-text-to-text` to more accurately reflect the multimodal nature of the tasks involving image-text prompts and textual/numerical output.
- Incorporating a "Sample Usage" section with `bash` code snippets for Supervised Fine-Tuning and Evaluation, directly sourced from the project's GitHub README, to provide practical guidance on how to use the dataset with the associated code.
These changes improve the …