# Unleashing Diffusion Transformers for Visual Correspondence by Modulating Massive Activations

arXiv: https://arxiv.org/abs/2505.18584 · HTML: https://mlsanity.com/paper/2505.18584/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models.
- Builds on 23 of its 80 references that are in the map.
- First posted on arXiv 24 May 2025.

## Placed in

- Diffusion models / [Image generation, text-to-image and editing](https://mlsanity.com/diffusion/image-generation-text-to-image-and-editing/)

## Builds on (most cited first, 23 of 23)

- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Layer Normalization](https://mlsanity.com/paper/1607.06450/) (2016)
- [GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models](https://mlsanity.com/paper/2112.10741/) (2021)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [The 2017 DAVIS Challenge on Video Object Segmentation](https://mlsanity.com/paper/1704.00675/) (2017)
- [Vision Transformers Need Registers](https://mlsanity.com/paper/2309.16588/) (2023)
- [PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis](https://mlsanity.com/paper/2310.00426/) (2023)
- [Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think](https://mlsanity.com/paper/2410.06940/) (2024)
- [CoTracker: It is Better to Track Together](https://mlsanity.com/paper/2307.07635/) (2023)
- [Label-Efficient Semantic Segmentation with Diffusion Models](https://mlsanity.com/paper/2112.03126/) (2021)
- [YouTube-VOS: A Large-Scale Video Object Segmentation Benchmark](https://mlsanity.com/paper/1809.03327/) (2018)
- [Deep ViT Features as Dense Visual Descriptors](https://mlsanity.com/paper/2112.05814/) (2021)
- [D2-Net: A Trainable CNN for Joint Detection and Description of Local Features](https://mlsanity.com/paper/1905.03561/) (2019)
- [Massive Activations in Large Language Models](https://mlsanity.com/paper/2402.17762/) (2024)
- [SegDiff: Image Segmentation with Diffusion Probabilistic Models](https://mlsanity.com/paper/2112.00390/) (2021)
- [Unsupervised Semantic Segmentation by Distilling Feature Correspondences](https://mlsanity.com/paper/2203.08414/) (2022)
- [SPair-71k: A Large-scale Benchmark for Semantic Correspondence](https://mlsanity.com/paper/1908.10543/) (2019)
- [Matcher: Segment Anything with One Shot Using All-Purpose Feature Matching](https://mlsanity.com/paper/2305.13310/) (2023)
- [AP-10K: A Benchmark for Animal Pose Estimation in the Wild](https://mlsanity.com/paper/2108.12617/) (2021)
- [Massive Values in Self-Attention Modules are the Key to Contextual Knowledge Understanding](https://mlsanity.com/paper/2502.01563/) (2025)
- [SLiMe: Segment Like Me](https://mlsanity.com/paper/2309.03179/) (2023)
- [Unveiling A Core Linguistic Region in Large Language Models](https://mlsanity.com/paper/2310.14928/) (2023)
- [Eye-for-an-eye: Appearance Transfer with Semantic Correspondence in Diffusion Models](https://mlsanity.com/paper/2406.07008/) (2024)

## Built on by (most cited first, 23 of 23)

- [Attention (as Discrete-Time Markov) Chains](https://mlsanity.com/paper/2507.17657/) (2025)
- [Hidden Dynamics of Massive Activations in Transformer Training](https://mlsanity.com/paper/2508.03616/) (2025)
- [Massive Activations are the Key to Local Detail Synthesis in Diffusion Transformers](https://mlsanity.com/paper/2510.11538/) (2025)
- [Probing and Leveraging Video Diffusion Transformer Features for Robust Point Tracking](https://mlsanity.com/paper/2512.20606/) (2025)
- [STRIDE: Training-Free Diversity Guidance via PCA-Directed Feature Perturbation in Single-Step Diffusion Models](https://mlsanity.com/paper/2605.11494/) (2026)
- [Steering Video Diffusion Transformers with Massive Activations](https://mlsanity.com/paper/2603.17825/) (2026)
- [Few Channels Draw The Whole Picture: Revealing Massive Activations in Diffusion Transformers](https://mlsanity.com/paper/2605.13974/) (2026)
- [Registers Matter for Pixel-Space Diffusion Transformers](https://mlsanity.com/paper/2605.16147/) (2026)
- [FreeStory: Training-Free Character Consistency for Free-Form Visual Storytelling](https://mlsanity.com/paper/2606.25079/) (2026)
- [Early Semantic Grounding in Image Editing Models for Zero-Shot Referring Image Segmentation](https://mlsanity.com/paper/2605.13122/) (2026)
- [Exploring Conditions for Diffusion models in Robotic Control](https://mlsanity.com/paper/2510.15510/) (2025)
- [Expanding the Role of Diffusion Models for Robust Classifier Training](https://mlsanity.com/paper/2602.19931/) (2026)
- [DA-Flow: Degradation-Aware Optical Flow Estimation with Diffusion Models](https://mlsanity.com/paper/2603.23499/) (2026)
- [Rethinking Where to Edit: Task-Aware Localization for Instruction-Based Image Editing](https://mlsanity.com/paper/2604.20258/) (2026)
- [SOCO: Benchmarking Semantic Object Correspondence in Vision Foundation Models](https://mlsanity.com/paper/2605.31597/) (2026)
- [MMDiff: Extending Diffusion Transformers for Multi-Modal Generation](https://mlsanity.com/paper/2606.16673/) (2026)
- [Awakening Diffusion Transformers: Eliciting Stronger Generation and Understanding via Massive Activation Modulation](https://mlsanity.com/paper/2607.02968/) (2026)
- [SPARK: Input-Conditioned Sparse Activation Modulation for Frozen DiT-based Super-Resolution](https://mlsanity.com/paper/2609.03813/) (2026)
- [ReaDiT Guidance: Control for Image and Video Generation using Diffusion Transformer Features](https://mlsanity.com/paper/2609.04649/) (2026)
- [Diffusion Trajectory Modeling for Semantic Correspondence](https://mlsanity.com/paper/2609.15357/) (2026)
- [Should We Skip Diffusion?](https://mlsanity.com/paper/2610.07002/) (2026)
- [Why DiT Models Underperform as Representation Learners without Long Skip Connections](https://mlsanity.com/paper/2610.13344/) (2026)
- [We Have Been Interpreting Diffusion Transformer Backbone Behaviors Wrong](https://mlsanity.com/paper/2610.14517/) (2026)

## Abstract (arXiv)

Pre-trained stable diffusion models (SD) have shown great advances in visual correspondence. In this paper, we investigate the capabilities of Diffusion Transformers (DiTs) for accurate dense correspondence. Distinct from SD, DiTs exhibit a critical phenomenon in which very few feature activations exhibit significantly larger values than others, known as textitmassive activations, leading to uninformative representations and significant performance degradation for DiTs. The massive activations consistently concentrate at very few fixed dimensions across all image patch tokens, holding little local information. We analyze these dimension-concentrated massive activations and uncover that their concentration is inherently linked to the Adaptive Layer Normalization (AdaLN) in DiTs. Building on these findings, we propose the Diffusion Transformer Feature (DiTF), a training-free AdaLN-based framework that extracts semantically discriminative features from DiTs. Specifically, DiTF leverages AdaLN to adaptively localize and normalize massive activations through channel-wise modulation. Furthermore, a channel discard strategy is introduced to mitigate the adverse effects of massive activations. Experimental results demonstrate that our DiTF outperforms both DINO and SD-based models and establishes a new state-of-the-art performance for DiTs in different visual correspondence tasks (eg, with +9.4% on Spair-71k and +4.4% on AP-10K-C.S.).