# Rethinking Multi-Condition DiTs: Eliminating Redundant Attention via Position-Alignment and Keyword-Scoping

arXiv: https://arxiv.org/abs/2602.06850 · HTML: https://mlsanity.com/paper/2602.06850/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 36 of its 41 references that are in the map.
- First posted on arXiv 6 Feb 2026.

## Builds on (most cited first, 30 of 36)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [Adding Conditional Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.05543/) (2023)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Scaling Rectified Flow Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2403.03206/) (2024)
- [IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Models](https://mlsanity.com/paper/2308.06721/) (2023)
- [InstantID: Zero-shot Identity-Preserving Generation in Seconds](https://mlsanity.com/paper/2401.07519/) (2024)
- [OminiControl: Minimal and Universal Control for Diffusion Transformer](https://mlsanity.com/paper/2411.15098/) (2024)
- [UniControl: A Unified Diffusion Model for Controllable Visual Generation In the Wild](https://mlsanity.com/paper/2305.11147/) (2023)
- [ControlNet++: Improving Conditional Controls with Efficient Consistency Feedback](https://mlsanity.com/paper/2404.07987/) (2024)
- [Learning-to-Cache: Accelerating Diffusion Transformer via Layer Caching](https://mlsanity.com/paper/2406.01733/) (2024)
- [Prodigy: An Expeditiously Adaptive Parameter-Free Learner](https://mlsanity.com/paper/2306.06101/) (2023)
- [EasyControl: Adding Efficient and Flexible Control for Diffusion Transformer](https://mlsanity.com/paper/2503.07027/) (2025)
- [Diffusion Models for Image Restoration and Enhancement: A Comprehensive Survey](https://mlsanity.com/paper/2308.09388/) (2023)
- [CreatiLayout: Siamese Multimodal Diffusion Transformer for Creative Layout-to-Image Generation](https://mlsanity.com/paper/2412.03859/) (2024)
- [OminiControl2: Efficient Conditioning for Diffusion Transformers](https://mlsanity.com/paper/2503.08280/) (2025)
- [TinyFusion: Diffusion Transformers Learned Shallow](https://mlsanity.com/paper/2412.01199/) (2024)
- [DiP-GO: A Diffusion Pruner via Few-step Gradient Optimization](https://mlsanity.com/paper/2410.16942/) (2024)
- [StyleAdapter: A Unified Stylized Image Generation Model](https://mlsanity.com/paper/2309.01770/) (2023)
- [Slide-Transformer: Hierarchical Vision Transformer with Local Self-Attention](https://mlsanity.com/paper/2304.04237/) (2023)
- [AutoStory: Generating Diverse Storytelling Images with Minimal Human Effort](https://mlsanity.com/paper/2311.11243/) (2023)
- [Ctrl-X: Controlling Structure and Appearance for Text-To-Image Generation Without Guidance](https://mlsanity.com/paper/2406.07540/) (2024)
- [UniCombine: Unified Multi-Conditional Combination with Diffusion Transformer](https://mlsanity.com/paper/2503.09277/) (2025)
- [DreamArtist++: Controllable One-Shot Text-to-Image Generation via Positive-Negative Adapter](https://mlsanity.com/paper/2211.11337/) (2022)
- [FastCache: Fast Caching for Diffusion Transformer Through Learnable Linear Approximation](https://mlsanity.com/paper/2505.20353/) (2025)
- [AnyStory: Towards Unified Single and Multiple Subject Personalization in Text-to-Image Generation](https://mlsanity.com/paper/2501.09503/) (2025)
- [Image Synthesis under Limited Data: A Survey and Taxonomy](https://mlsanity.com/paper/2307.16879/) (2023)

## Abstract (arXiv)

While modern text-to-image models excel at prompt-based generation, they often lack the fine-grained control necessary for specific user requirements like spatial layouts or subject appearances. Multi-condition control addresses this, yet its integration into Diffusion Transformers (DiTs) is bottlenecked by the conventional ``concatenate-and-attend'' strategy, which suffers from quadratic computational and memory overhead as the number of conditions scales. Our analysis reveals that much of this cross-modal interaction is spatially or semantically redundant. To this end, we propose Position-aligned and Keyword-scoped Attention (PKA), a highly efficient framework designed to eliminate these redundancies. Specifically, Position-Aligned Attention (PAA) linearizes spatial control by enforcing localized patch alignment, while Keyword-Scoped Attention (KSA) prunes irrelevant subject-driven interactions via semantic-aware masking. To facilitate efficient learning, we further introduce a Conditional Sensitivity-Aware Sampling (CSAS) strategy that reweights the training objective towards critical denoising phases, drastically accelerating convergence and enhancing conditional fidelity. Empirically, PKA delivers a 10.0× inference speedup and a 5.1× VRAM saving, providing a scalable and resource-friendly solution for high-fidelity multi-conditioned generation.