Cheers_edit

<div align="center">

Cheers: Decoupling Patch Details from Semantic Representations Enables Unified Multimodal Comprehension and Generation

Yichen Zhang^1*, Da Peng^2*, Zonghao Guo^1†, Zijian Zhang³, Xuesong Yang³,

Tong Sun³, Shichu Sun³, Yidan Zhang³, Yanghao Li¹, Haiyan Zhao¹, Wang Xu¹,

Qi Shi¹, Yangang Sun¹, Chi Chen¹, Shuo Wang¹, Yukun Yan¹, Xu Han¹,

Qiang Ma¹, Wei Ke², Liang Wang³, Zhiyuan Liu¹, Maosong Sun¹

¹Tsinghua University, ²Xi'an Jiaotong University, ³University of Chinese Academy of Sciences

* Equal contribution † Corresponding author

🌟 What is Cheers?

A recent cutting-edge topic in multimodal modeling is to unify visual comprehension and generation within a single model. However, the two tasks demand mismatched decoding regimes and visual representations, making it non-trivial to jointly optimize within a shared feature space. In this work, we present Cheers, a unified multimodal model that decouples patch-level details from semantic representations, thereby stabilizing semantics for multimodal understanding and improving fidelity for image generation via gated detail residuals. Cheers includes three key components: (i) a unified vision tokenizer that encodes and compresses image latent states into semantic tokens for efficient LLM conditioning, (ii) an LLM-based Transformer that unifies autoregressive decoding for text generation and diffusion decoding for image generation, and (iii) a cascaded flow matching head that decodes visual semantics first and then injects semantically gated detail residuals from the vision tokenizer to refine high-frequency content. Experiments on popular benchmarks demonstrate that Cheers matches or surpasses advanced UMMs in both visual understanding and generation. Notably, Cheers outperforms the Tar-1.5B on the popular benchmarks GenEval and MMBench, while requiring only 20% of the training cost, indicating effective and efficient (i.e., 4x token compression) unified multimodal modeling.

<!-- Provide the basic links for the model. -->

• Repository: [https://github.com/AI9Stars/Cheers]
• Paper: [https://arxiv.org/abs/2603.12793]

</div>

Model Evaluation

compare with cheers

🔥 News

• [2026/03/19] 🎉 Demo is now available on Hugging Face. Thanks to Prithiv Sakthi for setting it up!
• [2026/03/16] 📢 The Cheers paper is officially released.
• [2026/03/16] 🛠 We open-source the evaluation code and training pipeline. Our codebase is highly efficient: training on 3.8M samples takes only about two days on a single machine with 8×A100 GPUs.
• [2026/03/16] 📦 The model checkpoints of Cheers are now available.

Model Card Contact

For any questions or collaborations, feel free to contact us : )

<p align="left"> 📧 <a href="[email protected]">[email protected]</a>&nbsp&nbsp | &nbsp&nbsp 📧 <a href="[email protected]">[email protected]</a>&nbsp&nbsp | &nbsp&nbsp 📧 <a href="[email protected]">[email protected]</a>&nbsp&nbsp </p>

📖 Citation

If you find Cheers useful, please cite Cheers technical report using this BibTeX.

@article{zhang2026cheers,
  title={CHEERS: DECOUPLING PATCH DETAILS FROM SEMANTIC REPRESENTATIONS ENABLES UNIFIED MULTIMODAL COMPREHENSION AND GENERATION},
  author={Zhang, Yichen and Peng, Da and Guo, Zonghao and Zhang, Zijian and Yang, Xuesong and Sun, Tong and Sun, Shichu and Zhang, Yidan and Li, Yanghao and Zhao, Haiyan and others},
  journal={arXiv preprint arXiv:2603.12793},
  year={2026}
}