Matter that learns what matters
Abstract
Can a material learn what deserves its attention? We show that it can in a model optical network, in which action outcomes train a persistent record in each link. The record biases whether a source is consulted and sets how strongly it changes a stored answer, so useful sources gain attention and influence without being identified in advance. The network's experience thus shapes the information available for its future learning. Resonant readouts make this learning fragile: reinforcement can carry a record past resonance, where it persists but its influence weakens. We derive bounds within which relevant signals guide the answer, irrelevant signals are rejected and stored answers remain open to revision. These bounds guide a learning rule that limits reinforcement as records form; in simulations, it acquires useful records and keeps them functional. When relevance changes, an obsolete record can block its own replacement. A local state that strengthens negative writing after unsuccessful actions releases obsolete records and completes the learning sequence more often. How a memory is read should decide how it is written.
Related
- What if dark matter is composite of least interacting particles?
- Equivariant generative diffusion learns and generalizes the structural ensemble of amorphous oxides
- Ultrafast firing effects in Y-doped zirconia nanoparticles: Heating rate matters even under “Conventional” conditions
- Detector orientation matters: Consequences for gravitational-wave background detectability
- Full Three-dimensional Cell Geometry Matters for Epithelial Fluidization
BibTeX
@misc{berloff2026matter,
title = {{Matter that learns what matters}},
author = {Natalia G. Berloff},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.34428},
eprint = {2610.34428},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.34428},
}
RIS
TY - UNPB
TI - Matter that learns what matters
AU - Natalia G. Berloff
PY - 2026
DA - 2026/10/25
PB - arXiv
DO - 10.48550/arxiv.2610.34428
UR - https://arxiv.org/abs/2610.34428
AB - Can a material learn what deserves its attention? We show that it can in a model optical network, in which action outcomes train a persistent record in each link. The record biases whether a source is consulted and sets how strongly it changes a stored answer, so useful sources gain attention and influence without being identified in advance. The network's experience thus shapes the information available for its future learning. Resonant readouts make this learning fragile: reinforcement can carry a record past resonance, where it persists but its influence weakens. We derive bounds within which relevant signals guide the answer, irrelevant signals are rejected and stored answers remain open to revision. These bounds guide a learning rule that limits reinforcement as records form; in simulations, it acquires useful records and keeps them functional. When relevance changes, an obsolete record can block its own replacement. A local state that strengthens negative writing after unsuccessful actions releases obsolete records and completes the learning sequence more often. How a memory is read should decide how it is written.
ER -