Pith. sign in

REVIEW 2 cited by

Beyond the Edge of Stability via Two-step Gradient Updates

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 2206.04172 v3 pith:DKB7QKJR submitted 2022-06-08 cs.LG math.OCstat.ML

classification cs.LGmath.OCstat.ML
keywords gradientlearninglocaltwo-stepupdatesanalysisbeyondconvergence
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Gradient Descent (GD) is a powerful workhorse of modern machine learning thanks to its scalability and efficiency in high-dimensional spaces. Its ability to find local minimisers is only guaranteed for losses with Lipschitz gradients, where it can be seen as a `bona-fide' discretisation of an underlying gradient flow. Yet, many ML setups involving overparametrised models do not fall into this problem class, which has motivated research beyond the so-called ``Edge of Stability'' (EoS), where the step-size crosses the admissibility threshold inversely proportional to the Lipschitz constant above. Perhaps surprisingly, GD has been empirically observed to still converge regardless of local instability and oscillatory behavior. The incipient theoretical analysis of this phenomena has mainly focused in the overparametrised regime, where the effect of choosing a large learning rate may be associated to a `Sharpness-Minimisation' implicit regularisation within the manifold of minimisers, under appropriate asymptotic limits. In contrast, in this work we directly examine the conditions for such unstable convergence, focusing on simple, yet representative, learning problems, via analysis of two-step gradient updates. Specifically, we characterize a local condition involving third-order derivatives that guarantees existence and convergence to fixed points of the two-step updates, and leverage such property in a teacher-student setting, under population loss. Finally, starting from Matrix Factorization, we provide observations of period-2 orbit of GD in high-dimensional settings with intuition of its dynamics, along with exploration into more general settings.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. From Logistic Regression to the Perceptron Algorithm: Exploring Gradient Descent with Large Step Sizes

    cs.LG 2024-12 conditional novelty 7.0 of 10

    Logistic regression with gradient descent and infinite step size is the batch perceptron, and a normalized version achieves an n times better iteration complexity.

  2. Criteria and Bias of Parameterized Linear Regression under Edge of Stability Regime

    math.OC 2024-12 conditional novelty 6.0 of 10

    Under specific conditions, gradient descent converges in the unstable edge-of-stability regime for a quadratic loss on a depth-2 diagonal linear network, with a bias bound depending on step size and initialization.

Pith tools