# Sensory robustness through top-down feedback and neural stochasticity in recurrent vision models

arXiv: https://arxiv.org/abs/2508.07115 · HTML: https://mlsanity.com/paper/2508.07115/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 36 of its 110 references that are in the map.
- First posted on arXiv 9 Aug 2025.

## Builds on (most cited first, 30 of 36)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Explaining and Harnessing Adversarial Examples](https://mlsanity.com/paper/1412.6572/) (2014)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality](https://mlsanity.com/paper/1801.02613/) (2018)
- [Intrinsic dimension of data representations in deep neural networks](https://mlsanity.com/paper/1905.12784/) (2019)
- [Estimating the intrinsic dimension of datasets by a minimal neighborhood information](https://mlsanity.com/paper/1803.06992/) (2018)
- [Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks](https://mlsanity.com/paper/1701.04722/) (2017)
- [Deep Neural Networks Reveal a Gradient in the Complexity of Neural Representations across the Brain's Ventral Visual Pathway](https://mlsanity.com/paper/1411.6422/) (2014)
- [Toward Next-Generation Artificial Intelligence: Catalyzing the NeuroAI Revolution](https://mlsanity.com/paper/2210.08340/) (2022)
- [Predictive Coding: a Theoretical and Experimental Review](https://mlsanity.com/paper/2107.12979/) (2021)
- [Recurrence is required to capture the representational dynamics of the human visual system](https://mlsanity.com/paper/1903.05946/) (2019)
- [Gated Feedback Recurrent Neural Networks](https://mlsanity.com/paper/1502.02367/) (2015)
- [Neural population geometry: An approach for understanding biological and artificial neural networks](https://mlsanity.com/paper/2104.07059/) (2021)
- [Deep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognition](https://mlsanity.com/paper/1406.3284/) (2014)
- [The neuroconnectionist research programme](https://mlsanity.com/paper/2209.03718/) (2022)
- [Learning to Combine Top-Down and Bottom-Up Signals in Recurrent Neural Networks with Attention over Modules](https://mlsanity.com/paper/2006.16981/) (2020)
- [Brain-inspired Computational Intelligence via Predictive Coding](https://mlsanity.com/paper/2308.07870/) (2023)
- [Task-Driven Convolutional Recurrent Models of the Visual System](https://mlsanity.com/paper/1807.00053/) (2018)
- [Hybrid Predictive Coding: Inferring, Fast and Slow](https://mlsanity.com/paper/2204.02169/) (2022)
- [Deep Predictive Coding Network with Local Recurrent Processing for Object Recognition](https://mlsanity.com/paper/1805.07526/) (2018)
- [Deep Predictive Coding Network for Object Recognition](https://mlsanity.com/paper/1802.04762/) (2018)
- [Predictive Coding, Variational Autoencoders, and Biological Connections](https://mlsanity.com/paper/2011.07464/) (2020)
- [Predify: Augmenting deep neural networks with brain-inspired predictive coding dynamics](https://mlsanity.com/paper/2106.02749/) (2021)
- [On 1/n neural representation and robustness](https://mlsanity.com/paper/2012.04729/) (2020)
- [A Theoretical Framework for Inference and Learning in Predictive Coding Networks](https://mlsanity.com/paper/2207.12316/) (2022)
- [Neural Population Geometry Reveals the Role of Stochasticity in Robust Perception](https://mlsanity.com/paper/2111.06979/) (2021)
- [A Measure of the Complexity of Neural Representations based on Partial Information Decomposition](https://mlsanity.com/paper/2209.10438/) (2022)
- [How Robust are Deep Neural Networks?](https://mlsanity.com/paper/1804.11313/) (2018)
- [Active Predictive Coding: A Unified Neural Framework for Learning Hierarchical World Models for Perception and Planning](https://mlsanity.com/paper/2210.13461/) (2022)

## Abstract (arXiv)

Biological systems leverage top-down feedback for visual processing, yet most artificial vision models succeed in image classification using purely feedforward or recurrent architectures, calling into question the functional significance of descending cortical pathways. Here, we trained convolutional recurrent neural networks (ConvRNN) on image classification in the presence or absence of top-down feedback projections to elucidate the specific computational contributions of those feedback pathways. We found that ConvRNNs with top-down feedback exhibited remarkable speed-accuracy trade-off and robustness to noise perturbations and adversarial attacks, but only when they were trained with stochastic neural variability, simulated by randomly silencing single units via dropout. By performing detailed analyses to identify the reasons for such benefits, we observed that feedback information substantially shaped the representational geometry of the post-integration layer, combining the bottom-up and top-down streams, and this effect was amplified by dropout. Moreover, feedback signals coupled with dropout optimally constrained network activity onto a low-dimensional manifold and encoded object information more efficiently in out-of-distribution regimes, with top-down information stabilizing the representational dynamics at the population level. Together, these findings uncover a dual mechanism for resilient sensory coding. On the one hand, neural stochasticity prevents unit-level co-adaptation albeit at the cost of more chaotic dynamics. On the other hand, top-down feedback harnesses high-level information to stabilize network activity on compact low-dimensional manifolds.