Mapping the learning curves of deep learning networks
PLOS Computational Biology · 2025
DOI: 10.1371/journal.pcbi.1012286 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS Computational Biology, 2025. Verfasst von Yanru Jiang und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Communication, University of California, Los Angeles, Los Angeles, California, United States of America. Hauptthemen laut MEDLINE: Deep Learning, Learning Curve, Neural Networks, Computer. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 69 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
There is an important challenge in systematically interpreting the internal representations of deep neural networks (DNNs). Existing techniques are often less effective for non-tabular tasks, or they primarily focus on qualitative, ad-hoc interpretations of models. In response, this study introduces a cognitive science-inspired, multi-dimensional quantification and visualization approach that captures two temporal dimensions of model learning: the “information-processing trajectory” and the “developmental trajectory.” The former represents the influence of incoming signals on an agent’s decision-making, while the latter conceptualizes the gradual improvement in an agent’s performance throughout its lifespan. Tracking the learning curves of DNNs enables researchers to explicitly identify the model appropriateness of a given task, examine the properties of the underlying input signals, and assess the model’s alignment (or lack thereof) with human learning experiences. To illustrate this method, we conducted 750 runs of simulations on two temporal tasks: gesture detection and sentence classification, showcasing its applicability across different types of deep learning tasks. Using four descriptive metrics to quantify the mapped learning curves— start , end - start , max , t max —, we identified significant differences in learning patterns based on data sources and class distinctions (all p’s < .0001), the prominent role of spatial semantics in gesture learning, and larger information gains in language learning. We highlight three key insights gained from mapping learning curves: non-monotonic progress , pairwise comparisons , and domain distinctions . We reflect on the theoretical implications of this method for cognitive processing, language models and representations from multiple modalities.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Communication, University of California, Los Angeles, Los Angeles, California, United States of America.
ORCID-Profile: Yanru Jiang
Zitieren
APA
Jiang, Y., & Dale, R. (2025). Mapping the learning curves of deep learning networks. PLOS Computational Biology. https://doi.org/10.1371/journal.pcbi.1012286
Vancouver
Jiang Y, Dale R. Mapping the learning curves of deep learning networks. PLOS Computational Biology. 2025. doi:10.1371/journal.pcbi.1012286
BibTeX
@article{Jiang2025,
author = {Jiang, Yanru and Dale, Rick},
title = {Mapping the learning curves of deep learning networks},
journal = {PLOS Computational Biology},
year = {2025},
doi = {10.1371/journal.pcbi.1012286}
}
EUVIMED