# Locating and Editing Figure-Ground Organization in Vision Transformers

arXiv: https://arxiv.org/abs/2603.06407 · HTML: https://mlsanity.com/paper/2603.06407/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 25 references that are in the map.
- First posted on arXiv 6 Mar 2026.

## Builds on (most cited first, 16 of 16)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Zero-Shot Text-to-Image Generation](https://mlsanity.com/paper/2102.12092/) (2021)
- [BEiT: BERT Pre-Training of Image Transformers](https://mlsanity.com/paper/2106.08254/) (2021)
- [ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness](https://mlsanity.com/paper/1811.12231/) (2018)
- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [Steering Llama 2 via Contrastive Activation Addition](https://mlsanity.com/paper/2312.06681/) (2023)
- [Patchscopes: A Unifying Framework for Inspecting Hidden Representations of Language Models](https://mlsanity.com/paper/2401.06102/) (2024)
- [Are Convolutional Neural Networks or Transformers more like human vision?](https://mlsanity.com/paper/2105.07197/) (2021)
- [Characterizing Mechanisms for Factual Recall in Language Models](https://mlsanity.com/paper/2310.15910/) (2023)
- [Self-attention in Vision Transformers Performs Perceptual Grouping, Not Attention](https://mlsanity.com/paper/2303.01542/) (2023)
- [Mixed Evidence for Gestalt Grouping in Deep Neural Networks](https://mlsanity.com/paper/2203.07302/) (2022)
- [Understanding Deep Convolutional Networks through Gestalt Theory](https://mlsanity.com/paper/1810.08697/) (2018)
- [From Local Cues to Global Percepts: Emergent Gestalt Organization in Self-Supervised Vision Models](https://mlsanity.com/paper/2506.00718/) (2025)
- [Finding Closure: A Closer Look at the Gestalt Law of Closure in Convolutional Neural Networks](https://mlsanity.com/paper/2408.12460/) (2024)

## Abstract (arXiv)

Vision Transformers must resolve figure-ground organization by choosing between completions driven by local geometric evidence and those favored by global organizational priors, giving rise to a characteristic perceptual ambiguity. We aim to locate where the canonical Gestalt prior convexity is realized within the internal components of BEiT. Using a controlled perceptual conflict based on synthetic shapes of darts, we systematically mask regions that equally admit either a concave completion or a convex completion. We show that BEiT reliably favors convex completion under this competition. Projecting internal activations into the model's discrete visual codebook space via logit attribution reveals that this preference is governed by identifiable functional units within transformer substructures. Specifically, we find that figure-ground organization is ambiguous through early and intermediate layers and resolves abruptly in later layers. By decomposing the direct effect of attention heads, we identify head L0H9 acting as an early seed, introducing a weak bias toward convexity. Downscaling this single attention head shifts the distributional mass of the perceptual conflict across a continuous decision boundary, allowing concave evidence to guide completion.