Activation Saturation and Floquet Spectrum Collapse in Neural ODEs
We prove that activation saturation imposes a structural dynamical limitation on autonomous Neural ODEs $\dot{h}=f_θ(h)$ with saturating activations ($\tanh$, sigmoid, etc.): if $q$ hidden layers of the MLP $f_θ$ satisfy $|σ'|\leδ$ on a region~$U$, the input Jacobian is attenuated as $\norm{Df_θ(x)}\le C(U)$ (for activations with $\sup_{x}|σ'(x)|\le 1$, e.g.\ $\tanh$ and sigmoid, this reduces to $C_Wδ^q$), forcing every Floquet (Lyapunov) exponen along any $T$-periodic orbit $γ\subset U$ into the interval $[-C(U),\;C(U)]$. This is a collapse of the Floquet spectrum: as saturation deepens…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2604.00543
curl -X POST /api/v1/records/arxiv/2604.00543/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2604.00543/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'