On 1/n neural representation and robustness
This work is significant for researchers in neuroscience and machine learning interested in understanding the link between neural representation structure and network robustness, offering an incremental step in this understanding.
This paper investigates how the covariance spectrum of neural representations affects robustness to perturbations in both mouse V1 and artificial neural networks. The authors found that imposing a 1/n covariance spectrum, observed experimentally in mouse V1, on artificial neural networks makes them more robust to adversarial attacks.
Understanding the nature of representation in neural networks is a goal shared by neuroscience and machine learning. It is therefore exciting that both fields converge not only on shared questions but also on similar approaches. A pressing question in these areas is understanding how the structure of the representation used by neural networks affects both their generalization, and robustness to perturbations. In this work, we investigate the latter by juxtaposing experimental results regarding the covariance spectrum of neural representations in the mouse V1 (Stringer et al) with artificial neural networks. We use adversarial robustness to probe Stringer et al's theory regarding the causal role of a 1/n covariance spectrum. We empirically investigate the benefits such a neural code confers in neural networks, and illuminate its role in multi-layer architectures. Our results show that imposing the experimentally observed structure on artificial neural networks makes them more robust to adversarial attacks. Moreover, our findings complement the existing theory relating wide neural networks to kernel methods, by showing the role of intermediate representations.