Paper Detail
Ceyuan Yang, Zhijie Lin, Yang Zhao, Fei Xiao, Hao He, Qi Zhao, Chaorui Deng, Kunchang Li, Zihan Ding, Yuwei Guo, Fuyun Wang, Fangqi Zhu, Xiaonan Nie, Shenhan Zhu, Shanchuan Lin, Hongsheng Li, Weilin Huang, Guang Shi, Haoqi Fan
We present Omni, a unified multimodal model natively trained on diverse modalities, including text, images, videos, 3D geometry, and hidden representations. We find that such training enables Context Unrolling, where the model explicitly reasons across multiple modal representations before producing predictions. This process enables the model to aggregate complementary information across heterogeneous modalities, facilitating a more faithful approximation of the shared multimodal knowledge manifold and improving downstream reasoning fidelity. As a result, Omni achieves strong performance on both multimodal generation and understanding benchmarks, while demonstrating advanced multimodal reasoning capabilities, including in-context generation of text, image, video, and 3D geometry.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@article{yang2026context,
title = {Context Unrolling in Omni Models},
author = {Ceyuan Yang and Zhijie Lin and Yang Zhao and Fei Xiao and Hao He and Qi Zhao and Chaorui Deng and Kunchang Li and Zihan Ding and Yuwei Guo and Fuyun Wang and Fangqi Zhu and Xiaonan Nie and Shenhan Zhu and Shanchuan Lin and Hongsheng Li and Weilin Huang and Guang Shi and Haoqi Fan},
year = {2026},
abstract = {We present Omni, a unified multimodal model natively trained on diverse modalities, including text, images, videos, 3D geometry, and hidden representations. We find that such training enables Context Unrolling, where the model explicitly reasons across multiple modal representations before producing predictions. This process enables the model to aggregate complementary information across heterogeneous modalities, facilitating a more faithful approximation of the shared multimodal knowledge manif},
url = {https://arxiv.org/abs/2604.21921},
keywords = {cs.CV, multimodal model, context unrolling, multimodal knowledge manifold, downstream reasoning, in-context generation, huggingface daily},
eprint = {2604.21921},
archiveprefix = {arXiv},
}
{}