README.md

metadata

size_categories:
  - n<1K
task_categories:
  - question-answering
language:
  - en
tags:
  - agent
  - thinking-with-images

O3-Bench

@inproceedings{
  anonymous2025insighto3,
  title={InSight-o3: Empowering Multimodal Foundation Models with Generalized Visual Search},
  author={Anonymous},
  booktitle={Submitted to The Fourteenth International Conference on Learning Representations},
  year={2025},
  url={https://openreview.net/forum?id=vlraTIgUD3},
  note={under review}
}