Readout Representation: Redefining Neural Codes by Input Recovery

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Onoo, Shunsuke, Nagano, Yoshihiro, Kamitani, Yukiyasu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915553434664960
author Onoo, Shunsuke
Nagano, Yoshihiro
Kamitani, Yukiyasu
author_facet Onoo, Shunsuke
Nagano, Yoshihiro
Kamitani, Yukiyasu
contents Sensory representation is typically understood through a hierarchical-causal framework where progressively abstract features are extracted sequentially. However, this causal view fails to explain misrepresentation, a phenomenon better handled by an informational view based on decodable content. This creates a tension: how does a system that abstracts away details still preserve the fine-grained information needed for downstream functions? We propose readout representation to resolve this, defining representation by the information recoverable from features rather than their causal origin. Empirically, we show that inputs can be accurately reconstructed even from heavily perturbed mid-level features, demonstrating that a single input corresponds to a broad, redundant region of feature space, challenging the causal mapping perspective. To quantify this property, we introduce representation size, a metric linked to model robustness and representational redundancy. Our framework offers a new lens for analyzing how both biological and artificial neural systems learn complex features while maintaining robust, information-rich representations of the world.
format Preprint
id arxiv_https___arxiv_org_abs_2510_12228
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Readout Representation: Redefining Neural Codes by Input Recovery
Onoo, Shunsuke
Nagano, Yoshihiro
Kamitani, Yukiyasu
Neurons and Cognition
Sensory representation is typically understood through a hierarchical-causal framework where progressively abstract features are extracted sequentially. However, this causal view fails to explain misrepresentation, a phenomenon better handled by an informational view based on decodable content. This creates a tension: how does a system that abstracts away details still preserve the fine-grained information needed for downstream functions? We propose readout representation to resolve this, defining representation by the information recoverable from features rather than their causal origin. Empirically, we show that inputs can be accurately reconstructed even from heavily perturbed mid-level features, demonstrating that a single input corresponds to a broad, redundant region of feature space, challenging the causal mapping perspective. To quantify this property, we introduce representation size, a metric linked to model robustness and representational redundancy. Our framework offers a new lens for analyzing how both biological and artificial neural systems learn complex features while maintaining robust, information-rich representations of the world.
title Readout Representation: Redefining Neural Codes by Input Recovery
topic Neurons and Cognition
url https://arxiv.org/abs/2510.12228