# Vision-TTT: Efficient and Expressive Visual Representation Learning with Test-Time Training

arXiv: https://arxiv.org/abs/2603.00518 · HTML: https://mlsanity.com/paper/2603.00518/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Feb 2026.

## Built on by (most cited first, 2 of 2)

- [Representative Attention For Vision Transformers](https://mlsanity.com/paper/2605.14913/) (2026)
- [VisionHOPE: Visual Backbones as Self-Modifying Learning Systems](https://mlsanity.com/paper/2609.33325/) (2026)

## Abstract (arXiv)

Learning efficient and expressive visual representation has long been the pursuit of computer vision research. While Vision Transformers (ViTs) gradually replace traditional Convolutional Neural Networks (CNNs) as more scalable vision learners, their applications are plagued by the quadratic complexity of the self-attention mechanism. To address the challenge, we introduce a new linear-time sequence modeling method Test-Time Training (TTT) into vision and propose Vision-TTT, which treats visual sequences as datasets and compresses the visual token sequences in a novel self-supervised learning manner. By incorporating the dual-dataset strategy and Conv2d-based dataset preprocessing, Vision-TTT effectively extends vanilla TTT to model 2D visual correlations with global receptive fields. Extensive experiments show that textttVittt-T/S/B achieve 77.7%,81.8%,82.7% Top-1 accuracy on ImageNet classification and also greatly outperform their counterparts on downstream tasks. At 1280×1280 resolution, textttVittt-T reduces FLOPs by 79.4% and runs 4.72× faster with 88.9% less memory than DeiT-T. These results demonstrate the expressiveness and efficiency of Vision-TTT as a strong candidate for the next-generation generic visual backbone.