Sensory aligned receptive fields boost network accuracy depending on neuron complexity
The Computational Value of Sensory-Aligned Receptive Fields Depends on Neuronal Expressivity
Neural and Evolutionary ComputingMachine Learning
Summary
Our brains process sensory information through neurons that respond to specific features like sound frequency or visual position. This paper shows that artificial neural networks work better when their inputs are organized around meaningful sensory features relevant to the task. However, as individual neurons become more complex, this benefit decreases. Additionally, simply having sparse connections is not enough to match the performance gained by having these task-aligned receptive fields.
What this means in practice
- •For neural network developers: Design input layers with task-aligned receptive fields to improve accuracy on sensory classification tasks while controlling neuron complexity.
- •For neuromorphic engineers: Implement structured receptive fields in hardware neurons to efficiently enhance performance in event-based sensory processing systems.
Authors
Agnese Adorante, Aaron Spieler, Anna Levina
Abstract
Biological sensory neurons have selective receptive fields organized along meaningful stimulus coordinates, such as frequency, motion direction, or retinotopic position. Such structure may arise from efficient coding and biological constraints on activity, connectivity, and wiring, as computational studies of simple neurons have shown across modalities. This raises a question: do structured receptive fields confer a computational advantage beyond resource efficiency itself, and does this advantage persist when individual neurons are highly expressive? We address this question in recurrent networks of Expressive Leaky Memory neurons, where we can independently vary neuronal complexity and the organization of feed-forward receptive fields. Across auditory and event-based visual classification tasks, receptive fields aligned with a task-relevant sensory coordinate improve test accuracy relative to budget-matched random receptive fields. This advantage disappears when sensory coordinates are scrambled, or when receptive fields follow task-irrelevant coordinates, showing that the benefit comes from alignment with task geometry rather than restricted connectivity alone. Increasing neuronal complexity reduces the performance advantage of structured receptive fields. Finally, generic synaptic sparsity regularization induces input selectivity and partially recovers performance, but remains substantially below explicitly structured receptive fields, suggesting that sparsity alone is insufficient to recover the full computational benefit of task-aligned receptive fields. Together, our results show that appropriate receptive fields can serve as a computational prior beyond sparsity itself, and that their value depends on the computational expressivity of individual neurons.