Reference library
Self-classifying MNIST Digits
Paper summary
Trains neural cellular automata so local pixel-like cells communicate and reach a distributed classification of digit shape. Demonstrations examine adaptation to drawing changes and robustness under selected distortions.
Relevance to the project
A regular local-processing reference for global coordination and causal damage tests. The classification labels and training loss remain predefined objectives.
Distributed classification is not evidence of open-ended evolution.
Source record
- Reference ID
- article-https-distill-pub-2020-selforg-mnist
- Archived version
- Cited online article
- Summary provenance
- reference-library/papers/article-https-distill-pub-2020-selforg-mnist/metadata.json · summary
- Review status
- initial orientation summary; not a critical full-paper review
- Reading scope
- Selected abstract/opening and available closing discussion/conclusion passages in the preserved text extract; broader project notes provide context. No reproduction of experiments, exhaustive methods review, or equation/figure audit.