A General Close-loop Predictive Coding Framework for Auditory Working Memory

2025-03-16

AI Summary

Modeling auditory working memory remains a fundamental challenge in computational neuroscience and cognitive AI. Method: We propose the first general closed-loop predictive coding framework specifically designed for this task. Our approach employs a hierarchical closed-loop architecture with self-supervised temporal reconstruction objectives and cross-layer error feedback mechanisms, explicitly modeling the transient storage and dynamic updating of short-term acoustic signals. Contribution/Results: This work pioneers the systematic application of closed-loop predictive coding to auditory working memory, enabling unified modeling of both environmental sound and speech tasks for the first time. Evaluated on two major benchmark datasets, our method significantly outperforms RNN- and Transformer-based baselines, achieving superior semantic consistency and memory fidelity. These results empirically validate the effectiveness and generalizability of predictive coding for neuro-symbolic auditory cognitive modeling.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Neuro-Symbolic Learning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Large language models for search
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Auditory working memory is essential for various daily activities, such as language acquisition, conversation. It involves the temporary storage and manipulation of information that is no longer present in the environment. While extensively studied in neuroscience and cognitive science, research on its modeling within neural networks remains limited. To address this gap, we propose a general framework based on a close-loop predictive coding paradigm to perform short auditory signal memory tasks. The framework is evaluated on two widely used benchmark datasets for environmental sound and speech, demonstrating high semantic similarity across both datasets.

Problem

• Modeling auditory working memory in neural networks
• Addressing limited research on memory modeling
• Proposing a predictive coding framework for auditory tasks

Innovation

• Close-loop predictive coding framework
• Short auditory signal memory tasks
• High semantic similarity across datasets