Downloads · 30 days
13
11% of all-time downloads
RvR-7B-MoT
JiayiGuo821/RvR-7B-MoT
RvR-7B-MoT is a machine learning model from JiayiGuo821. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
<a href="https://arxiv.org/pdf/2604.25636" <img src="https://img.shields.io/badge/RvR-Paper-red?logo=arxiv&logoColor=red" alt="RvR Paper on arXiv" / </a <a href="https://github.com/LeapLabTHU/RvR" <img src="https://im…
All-time downloads
118
Public
Parameters
14.7B
58.8 GB on disk
Likes
3
Public
Repo makeup
Click a slice to open those files.
.safetensors58.8 GB · 100%
Parameter types
How the weights are stored.
BF1614.6B · 99%
Model card
From the Hugging Face model README
<a href="https://arxiv.org/pdf/2604.25636">
<img
src="https://img.shields.io/badge/RvR-Paper-red?logo=arxiv&logoColor=red"
alt="RvR Paper on arXiv"
/>
</a>
<a href="https://github.com/LeapLabTHU/RvR">
<img
src="https://img.shields.io/badge/RvR-Code-black?logo=github&logoColor=white"
alt="RvR GitHub"
/>
</a>
Enlarging Modification Space Boosts Image Refinement in Unified Multimodal Models
Jiayi Guo, Linqing Wang, Jiangshan Wang, Yang Yue, Zeyu Liu, Zhiyuan Zhao, Qinglin Lu,
Gao Huang, Chunyu Wang ✉️Tsinghua University · Tencent Hunyuan (HY)
We present Refinement via Regeneration (RvR), a novel framework that reformulates image refinement in unified multimodal models from an editing-based paradigm to a regeneration-based one. Instead of relying on intermediate editing instructions and enforcing pixel-level consistency, our method directly regenerates images conditioned on the target prompt and semantic representations of the initial image, thereby enlarging the effective modification space. This design enables more complete semantic alignment and avoids error accumulation from coarse instructions, leading to more flexible and accurate refinement.
- Created
- Apr 22, 2026
- Updated
- Apr 29, 2026