# Inference-Path Optimization via Circuit Duplication in Frozen Visual Transformers for Marine Species Classification

arXiv: https://arxiv.org/abs/2604.03428 · HTML: https://mlsanity.com/paper/2604.03428/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Interpreting vision and multimodal models, one of 13 directions in Mechanistic interpretability.
- Builds on 11 of its 21 references that are in the map.
- First posted on arXiv 3 Apr 2026.

## Placed in

- Mechanistic interpretability / [Interpreting vision and multimodal models](https://mlsanity.com/mechinterp/interpreting-vision-and-multimodal-models/)

## Builds on (most cited first, 11 of 11)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA](https://mlsanity.com/paper/2410.20672/) (2024)
- [Computer Vision and Deep Learning for Fish Classification in Underwater Habitats: A Survey](https://mlsanity.com/paper/2203.06951/) (2022)
- [AQUA20: A Benchmark Dataset for Underwater Species Classification under Challenging Conditions](https://mlsanity.com/paper/2506.17455/) (2025)
- [Multi-Label Plant Species Classification with Self-Supervised Vision Transformers](https://mlsanity.com/paper/2407.06298/) (2024)
- [Unlocking the potential of deep learning for marine ecology: overview, applications, and outlook](https://mlsanity.com/paper/2109.14737/) (2021)
- [Vision Transformers for Zero-Shot Clustering of Animal Images: A Comparative Benchmarking Study](https://mlsanity.com/paper/2602.03894/) (2026)
- [Mitigating Domain Drift in Multi Species Segmentation with DINOv2: A Cross-Domain Evaluation in Herbicide Research Trials](https://mlsanity.com/paper/2508.07514/) (2025)

## Abstract (arXiv)

Automated underwater species classification is constrained by annotation cost and environmental variation that limits the transferability of fully supervised models. Recent work has shown that frozen embeddings from self-supervised vision foundation models already provide a strong label-efficient baseline for marine image classification. Here we investigate whether this frozen-embedding regime can be improved at inference time, without fine-tuning or changing model weights. We apply Circuit Duplication, an inference-time method originally proposed for Large Language Models, in which a selected range of transformer layers is traversed twice during the forward pass. We evaluate on the class-imbalanced AQUA20 benchmark using frozen DINOv3 embeddings under two settings: global circuit selection, where a single duplicated circuit is chosen for the full dataset, and class-specific circuit selection, where each species may receive a different optimal circuit. Both settings use simple semi-supervised downstream classifiers. Circuit Duplication consistently improves over the standard frozen forward pass. At the maximum label budget, class-specific selection reaches a macro F1 of 0.875, closing the gap to the fully supervised ConvNeXt benchmark (0.889) to 1.4 points without any gradient-based training. Four species exceed their fully supervised reference, with octopus improving by +12.1 F1 points. Across all budgets, roughly 75% of classes prefer a class-specific circuit, indicating a genuinely class-dependent benefit. To our knowledge, this is the first application of Circuit Duplication to computer vision.