Meta-autoencoders: An approach to discovery and representation of relationships between dynamically evolving classes

2025-07-12

AI Summary

This paper addresses the challenge of modeling dynamic evolutionary relationships among categories—such as biological species—in natural evolution. We propose the Meta-Autoencoder (MAE), the first higher-order representation framework that treats the *family of autoencoders itself* as the modeling object. MAE employs self-supervised deep learning to jointly encode both the parameters and architectures of multiple category-specific autoencoders, explicitly capturing their phylogenetic derivations and shared homologous features. We provide a formal definition of MAE and construct concrete instances, empirically validating its effectiveness in representing evolutionary relationships and uncovering latent taxonomic structures. This work pioneers the integration of meta-learning principles into category-relation modeling, establishing a methodological bridge between interpretable evolutionary biology modeling and deep generative modeling.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Metareasoning and Metaheuristics
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Models for Web evolution
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

An autoencoder (AE) is a neural network that, using self-supervised training, learns a succinct parameterized representation, and a corresponding encoding and decoding process, for all instances in a given class. Here, we introduce the concept of a meta-autoencoder (MAE): an AE for a collection of autoencoders. Given a family of classes that differ from each other by the values of some parameters, and a trained AE for each class, an MAE for the family is a neural net that has learned a compact representation and associated encoder and decoder for the class-specific AEs. One application of this general concept is in research and modeling of natural evolution -- capturing the defining and the distinguishing properties across multiple species that are dynamically evolving from each other and from common ancestors. In this interim report we provide a constructive definition of MAEs, initial examples, and the motivating research directions in machine learning and biology.

Problem

• Discover relationships between dynamically evolving classes
• Represent multiple species' defining and distinguishing properties
• Learn compact encodings for families of autoencoders

Innovation

• Meta-autoencoder for multiple autoencoders representation
• Compact neural net for class-specific AEs encoding
• Application in modeling dynamically evolving species