Metraipolis › archive › dc_93182dd76d56

Theoretical Guarantees for One-Shot Magnitude Pruning and Compute-Adaptive Early Exit — arXiv:2609.12337

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.12337
License (as archived)
CC0-1.0
Archived
2026-10-16T03:15:37.830Z
SHA-256 of the archived text
fbc6782526c1c9a962e20363b307adf47d93dd55c1bfee3b5ed045f49265ea65
Stable URL of this page
https://metraipolis.com/documento/dc_93182dd76d56

How to cite

Theoretical Guarantees for One-Shot Magnitude Pruning and Compute-Adaptive Early Exit — arXiv:2609.12337 — arXiv (en), https://arxiv.org/abs/2609.12337 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_93182dd76d56 (Metraipolis archive dc_93182dd76d56, sha256 fbc6782526c1c9a962e20363b307adf47d93dd55c1bfee3b5ed045f49265ea65)
BibTeX
@misc{dc93182dd76d56,
  title = {Theoretical Guarantees for One-Shot Magnitude Pruning and Compute-Adaptive Early Exit — arXiv:2609.12337},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.12337},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_93182dd76d56; sha256 fbc6782526c1c9a962e20363b307adf47d93dd55c1bfee3b5ed045f49265ea65},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.12337. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We study compute reduction in neural networks through a unified partial versus full computation view, captured by one-shot magnitude pruning in the static regime and early exit in the adaptive regime. In an asymptotic single-neuron model, we prove a concentration theorem for one-shot magnitude pruning with explicit rates. We also introduce the conditional perceptron for early exit and show that its excess generalization error decays as a power of the compute gap, with an exponent that grows to infinity as the alignment between partial and full computations tends to one. We then extend the analysis to deep networks, characterizing how pruning-induced distortions accumulate with depth and deriving a corresponding compute-accuracy tradeoff for frozen-backbone early exit under a neural network Gaussian process model. Numerical simulations corroborate the predicted scaling laws.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-14 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.12337 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T03:05:44.035Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).