Paper Detail

SSR-Merge: Subspace Signal Routing for Training-Free LoRA Merging in Diffusion Models

Zhengxuan Wei, Yi Dong, Zonghui Li, Xianhui Lin, Xing Liu, Hong Gu, Shaofeng Zhang, Wenbin Li, Qi Fan

arxiv Score 7.3

Published 2026-06-09 · First seen 2026-06-10

General AI

Abstract

Low-Rank Adaptation (LoRA) merging can efficiently combine diverse generative capabilities from multiple trained LoRAs for a diffusion model. However, existing LoRA merging techniques often suffer from severe parameter interference, causing destructive collisions in the shared parameter space. To address this, we propose Subspace Signal Routing (SSR), which resolves interference by routing internal signals instead of performing parameter-space merge. Specifically, SSR first constructs a unified subspace by concatenating candidate LoRAs along the rank dimension. Next, SSR employs an inverse correlation matrix to decorrelate mixed signals within this space. Finally, a directional guide matrix steers these purified signals into their respective task-specific subspaces. We provide a rigorous theoretical analysis proving that SSR aligns with the Ordinary Least Squares (OLS) solution, thereby ensuring mathematical optimality. We utilize the additivity of sufficient statistics to design a streaming algorithm. This enables on-the-fly updates that significantly reduce memory overhead and computation time. Extensive experiments validate that SSR significantly outperforms state-of-the-art methods while maintaining comparable efficiency. Code is available at https://github.com/nagara214/SSR-Merge.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{wei2026ssr,
  title = {SSR-Merge: Subspace Signal Routing for Training-Free LoRA Merging in Diffusion Models},
  author = {Zhengxuan Wei and Yi Dong and Zonghui Li and Xianhui Lin and Xing Liu and Hong Gu and Shaofeng Zhang and Wenbin Li and Qi Fan},
  year = {2026},
  abstract = {Low-Rank Adaptation (LoRA) merging can efficiently combine diverse generative capabilities from multiple trained LoRAs for a diffusion model. However, existing LoRA merging techniques often suffer from severe parameter interference, causing destructive collisions in the shared parameter space. To address this, we propose Subspace Signal Routing (SSR), which resolves interference by routing internal signals instead of performing parameter-space merge. Specifically, SSR first constructs a unified },
  url = {https://arxiv.org/abs/2606.10617},
  keywords = {cs.CV},
  eprint = {2606.10617},
  archiveprefix = {arXiv},
}

Metadata

{}