# RefRoute: Decoupling Conditioning Cost from References via Compact Residual Conditioning and Spatial Routing

arXiv: https://arxiv.org/abs/2610.07720 · HTML: https://mlsanity.com/paper/2610.07720/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 43 references that are in the map.
- First posted on arXiv 6 Oct 2026.

## Builds on (most cited first, 11 of 11)

- [Qwen-Image Technical Report](https://mlsanity.com/paper/2508.02324/) (2025)
- [FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space](https://mlsanity.com/paper/2506.15742/) (2025)
- [Emerging Properties in Unified Multimodal Pretraining](https://mlsanity.com/paper/2505.14683/) (2025)
- [In-Context LoRA for Diffusion Transformers](https://mlsanity.com/paper/2410.23775/) (2024)
- [Echo-4o: Harnessing the Power of GPT-4o Synthetic Images for Improved Image Generation](https://mlsanity.com/paper/2508.09987/) (2025)
- [MultiBind: A Benchmark for Attribute Misbinding in Multi-Subject Generation](https://mlsanity.com/paper/2603.21937/) (2026)
- [MACRO: Advancing Multi-Reference Image Generation with Structured Long-Context Data](https://mlsanity.com/paper/2603.25319/) (2026)
- [Three-Body Scattering for Generative Modeling](https://mlsanity.com/paper/2607.18198/) (2026)
- [NanoControl: A Lightweight Framework for Precise and Efficient Control in Diffusion Transformer](https://mlsanity.com/paper/2508.10424/) (2025)
- [UniCustom: Unified Visual Conditioning for Multi-Reference Image Generation](https://mlsanity.com/paper/2605.12088/) (2026)
- [StructGen: Disambiguating Multi-Reference Image Generation via Structured Context Modeling](https://mlsanity.com/paper/2607.15619/) (2026)

## Abstract (arXiv)

Multi-reference image generation requires preserving the appearance of multiple subjects while composing them into a coherent scene. However, existing diffusion transformers commonly encode references as dense visual token grids and jointly process them with global attention, making conditioning increasingly expensive as the number and resolution of references grow. We present RefRoute, a framework that addresses both reference representation cost and attention overhead through two complementary mechanisms. Compact residual conditioning combines low-resolution latent tokens with lightweight residual features extracted from full-resolution pixels, reducing reference token counts while retaining fine-grained appearance cues. Condition routing and attention routing align reference tokens with their assigned target regions and restrict cross-reference interactions, while allowing selective reference access beyond region boundaries for scene integration. We further introduce RefRoute-Data for training many-reference generation models and ManyRef100, a benchmark spanning human, object, and mixed compositions with 10-17 references. After many-reference fine-tuning, RefRoute achieves an overall Weighted-Ref-VIEScore of 36.06 on ManyRef100, compared with 8.88 for FLUX.2-Klein-9B. Separate inference-cost evaluations show substantially slower latency growth as the reference count increases: at 16 references, our 50-step and 4-step configurations achieve 18.3× and 14.2× speedups over their corresponding FLUX baselines, respectively. These results establish compact reference representations and spatially routed attention as an effective approach to scalable many-reference image generation.