# Fake & Square: Training Self-Supervised Vision Transformers with Synthetic Data and Synthetic Hard Negatives

arXiv: https://arxiv.org/abs/2509.02029 · HTML: https://mlsanity.com/paper/2509.02029/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 29 references that are in the map.
- First posted on arXiv 2 Sep 2025.

## Builds on (most cited first, 22 of 22)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [A Style-Based Generator Architecture for Generative Adversarial Networks](https://mlsanity.com/paper/1812.04948/) (2018)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [Unsupervised Learning of Visual Features by Contrasting Cluster Assignments](https://mlsanity.com/paper/2006.09882/) (2020)
- [An Empirical Study of Training Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.02057/) (2021)
- [Deep Clustering for Unsupervised Learning of Visual Features](https://mlsanity.com/paper/1807.05520/) (2018)
- [Self-supervised Learning: Generative or Contrastive](https://mlsanity.com/paper/2006.08218/) (2020)
- [A Theoretical Analysis of Contrastive Unsupervised Representation Learning](https://mlsanity.com/paper/1902.09229/) (2019)
- [Hard Negative Mixing for Contrastive Learning](https://mlsanity.com/paper/2010.01028/) (2020)
- [Fake it till you make it: Learning transferable representations from synthetic ImageNet clones](https://mlsanity.com/paper/2212.08420/) (2022)
- [Self-Supervised Learning with Swin Transformers](https://mlsanity.com/paper/2105.04553/) (2021)
- [Synthetic Data for Deep Learning](https://mlsanity.com/paper/1909.11512/) (2019)
- [Relay Diffusion: Unifying diffusion process across resolutions for image synthesis](https://mlsanity.com/paper/2309.03350/) (2023)
- [Learning Vision from Models Rivals Learning Vision from Data](https://mlsanity.com/paper/2312.17742/) (2023)
- [This dataset does not exist: training models from generated images](https://mlsanity.com/paper/1911.02888/) (2019)
- [Understanding Hard Negatives in Noise Contrastive Estimation](https://mlsanity.com/paper/2104.06245/) (2021)

## Abstract (arXiv)

This paper does not introduce a new method per se. Instead, we build on existing self-supervised learning approaches for vision, drawing inspiration from the adage "fake it till you make it". While contrastive self-supervised learning has achieved remarkable success, it typically relies on vast amounts of real-world data and carefully curated hard negatives. To explore alternatives to these requirements, we investigate two forms of "faking it" in vision transformers. First, we study the potential of generative models for unsupervised representation learning, leveraging synthetic data to augment sample diversity. Second, we examine the feasibility of generating synthetic hard negatives in the representation space, creating diverse and challenging contrasts. Our framework - dubbed Syn2Co - combines both approaches and evaluates whether synthetically enhanced training can lead to more robust and transferable visual representations on DeiT-S and Swin-T architectures. Our findings highlight the promise and limitations of synthetic data in self-supervised learning, offering insights for future work in this direction.