Landscape analysis for shallow neural networks: Complete classification of critical points for cubic activation and affine target functions
Source: arxiv.org/abs/2607.15173
In this paper, we study the optimization landscape induced by the true loss for shallow polynomial neural networks (PNNs) with $\mathfrak{h} \in \mathbb{N}$ neurons on the hidden layer, one-dimensional input and output layers, and a monomial activation of degree $d \in \mathbb{N}$, trained against a non-constant affine linear target function. Our first main result provides for arbitrary activation degree $d$ a sharp existence/non-existence criterion for \emph{global minimizers} with necessary structural conditions. We show that the infimum of the loss is always zero and achievable with at…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2607.15173
curl -X POST /api/v1/records/arxiv/2607.15173/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2607.15173/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'