Neural networks are compressed by lumping neurons with approximately matching dynamics in a polynomial ODE encoding, yielding substantial size reduction with preserved accuracy on synthetic and regression tasks.
arXiv preprint arXiv:2012.09243 (2020)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Neural Network Compression by Approximate Differential Equivalence
Neural networks are compressed by lumping neurons with approximately matching dynamics in a polynomial ODE encoding, yielding substantial size reduction with preserved accuracy on synthetic and regression tasks.