Pith. sign in

REVIEW 13 cited by

Understanding Deep Neural Networks with Rectified Linear Units

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 1611.01491 v6 pith:A23CZDWC submitted 2016-11-04 cs.LG cond-mat.dis-nncs.AIcs.CCstat.ML

Understanding Deep Neural Networks with Rectified Linear Units

classification cs.LG cond-mat.dis-nncs.AIcs.CCstat.ML
keywords reludeepexponentialfamilyfunctionshiddensizeconstruction
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

In this paper we investigate the family of functions representable by deep neural networks (DNN) with rectified linear units (ReLU). We give an algorithm to train a ReLU DNN with one hidden layer to *global optimality* with runtime polynomial in the data size albeit exponential in the input dimension. Further, we improve on the known lower bounds on size (from exponential to super exponential) for approximating a ReLU deep net function by a shallower ReLU net. Our gap theorems hold for smoothly parametrized families of "hard" functions, contrary to countable, discrete families known in the literature. An example consequence of our gap theorems is the following: for every natural number $k$ there exists a function representable by a ReLU DNN with $k^2$ hidden layers and total size $k^3$, such that any ReLU DNN with at most $k$ hidden layers will require at least $\frac{1}{2}k^{k+1}-1$ total nodes. Finally, for the family of $\mathbb{R}^n\to \mathbb{R}$ DNNs with ReLU activations, we show a new lowerbound on the number of affine pieces, which is larger than previous constructions in certain regimes of the network architecture and most distinctively our lowerbound is demonstrated by an explicit construction of a *smoothly parameterized* family of functions attaining this scaling. Our construction utilizes the theory of zonotopes from polyhedral theory.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 13 Pith papers

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

  1. Reconstructing the Stripping History of the Sagittarius Stream with Neural Networks

    astro-ph.GA 2026-05 unverdicted novelty 7.0

    A neural network trained on simulations infers stripping times for Sagittarius stream stars from phase-space data, measuring a 0.3 dex/Gyr metallicity gradient and estimating ages for globular clusters such as Pal 12 ...

  2. Non-Uniqueness of Solutions in Neural Variational Methods

    math.NA 2026-05 unverdicted novelty 7.0

    Finite linear measurements in variational neural discretizations cause ill-posed discrete problems with non-unique minimizers, independent of the underlying continuous variational problem's well-posedness.

  3. Key and Value Weights Are Probably All You Need: On the Necessity of the Query, Key, Value weight Triplet in Self-Attention Transformers

    cs.LG 2025-10 unverdicted novelty 7.0

    One of the Q, K or V weights in transformer self-attention is redundant and replaceable by the identity matrix under mild assumptions, reducing parameters by 25 percent with no loss in small-model performance.

  4. Estimation of High Dimensional Bounded Discrete Graphical Models via Regularized Generalized Score Matching

    stat.ME 2026-06 unverdicted novelty 6.0

    Introduces bounded discrete graphical models and the BRIDGE regularized score matching estimator with nonasymptotic error bounds and exact support recovery for high-dimensional discrete data.

  5. A Theory on Flow Matching with Neural Networks

    cs.LG 2026-06 unverdicted novelty 6.0

    Establishes convergence guarantees for overparameterized 2-layer ReLU networks in flow matching, generalization bounds for the velocity-field objective, and Wasserstein guarantees for generated samples, using multi-ta...

  6. Non-Uniqueness of Solutions in Neural Variational Methods

    math.NA 2026-05 unverdicted novelty 6.0

    Variational neural discretizations are structurally ill-posed with non-unique minimizers due to finite linear measurements, independent of the continuous variational problem's well-posedness.

  7. "Show Me You Comply... Without Showing Me Anything": Zero-Knowledge Software Auditing for AI-Enabled Systems

    cs.SE 2025-10 unverdicted novelty 6.0

    ZKMLOps is an MLOps framework that uses zero-knowledge proofs to generate verifiable cryptographic evidence of AI model compliance without revealing confidential information.

  8. Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models

    cs.LG 2024-01 unverdicted novelty 6.0

    SPIN lets weak LLMs become strong by self-generating training data from previous model versions and training to prefer human-annotated responses over its own outputs, outperforming DPO even with extra GPT-4 data on be...

  9. Low-Cost Stereo Vision for Robust 3D Positioning of Thin Radiata Pine Branches in Autonomous Drone Pruning

    cs.CV 2026-05 unverdicted novelty 5.0

    A drone-mounted stereo camera pipeline with YOLO segmentation, deep stereo depth, centroid triangulation, and MAD outlier rejection achieves robust 3D positioning of thin pine branches at 1-2 m distances.

  10. On Symmetry and Initialization for Neural Networks

    cs.LG 2019-07 unverdicted novelty 5.0

    For symmetric target functions, chosen initial conditions in one-hidden-layer networks enable SGD to produce generalization guarantees, unlike random initialization.

  11. Bayesian meta-learning for modeling Alzheimer's disease progression

    stat.ML 2026-06 unverdicted novelty 4.0

    Bayesian meta-learner predicts individualized Alzheimer's disease progression distributions from MRI and trajectories, competitive on ADNI data and less overconfident for long-term scores than deterministic versions.

  12. Positioning radiata pine branches requiring pruning by drone stereo vision

    cs.CV 2026-04 unverdicted novelty 3.0

    Drone stereo vision pipeline segments pine branches with YOLO variants and estimates depth with deep stereo networks, yielding more coherent maps than SGBM at 1-2 m distances.

  13. Deep learning applied to computational mechanics: A comprehensive review, state of the art, and the classics

    cs.LG 2022-12 unverdicted novelty 2.0

    A comprehensive review of deep learning techniques for computational mechanics, including LSTM for constitutive modeling, PINNs for PDE solving, optimizers, and kernel methods.