Pith. sign in

REVIEW 4 cited by

Benefits of depth in neural networks

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1602.04485 v2 pith:QDD4N4OF submitted 2016-02-14 cs.LG cs.NEstat.ML

classification cs.LGcs.NEstat.ML
keywords networksnodesgatesreluthetabenefitsdepthlayers
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
abstract

For any positive integer $k$, there exist neural networks with $\Theta(k^3)$ layers, $\Theta(1)$ nodes per layer, and $\Theta(1)$ distinct parameters which can not be approximated by networks with $\mathcal{O}(k)$ layers unless they are exponentially large --- they must possess $\Omega(2^k)$ nodes. This result is proved here for a class of nodes termed "semi-algebraic gates" which includes the common choices of ReLU, maximum, indicator, and piecewise polynomial functions, therefore establishing benefits of depth against not just standard networks with ReLU gates, but also convolutional networks with ReLU and maximization gates, sum-product networks, and boosted decision trees (in this last case with a stronger separation: $\Omega(2^{k^3})$ total tree nodes are required).

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. On Universality of Deep Equivariant Networks

    stat.ML 2025-10 conditional novelty 7.0 of 10

    Deep equivariant networks are universal over the entry-wise separable regime once depth stabilizes separation or a convolutional readout is added, unifying prior architecture-specific results.

  2. Algebraic Representability as the Limiting Regime of Grokking: An Exactly Solvable Model with Holomorphic Activations

    cs.LG 2026-07 conditional novelty 6.0 of 10

    A task ma+nb mod p is representable by a z^k holomorphic network iff m+n=k; non-representable tasks cannot be memorised at any width.

  3. Discrete Functional Geometry of ReLU Networks via ReLU Transition Graphs

    cs.LG 2025-09 reject novelty 4.0 of 10

    The paper claims ReLU Transition Graphs of ReLU networks are expanders whose spectral gap, region entropy, and edge KL divergence bound generalization and capacity; the proofs are sketches, and the empirical checks are weak.

  4. Iterate to Accelerate: A Unified Framework for Iterative Reasoning and Feedback Convergence

    cs.LG 2025-02 reject novelty 4.0 of 10

    The paper claims a unified O(1/t^2) acceleration theorem and a feedback-versus-feedforward separation, but the proof steps are invalid and the main results are false as stated.

Pith tools