Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

Last-Iterate Complexity of SGD for Convex and Smooth Stochastic Problems

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read SGD's last iterate reaches O(ln T/√T) for convex smooth problems, with no uniform variance assumption.

desk verdict Solid last-iterate result that settles an open question; printed proof has a fixable sign-error in Lemma 4.3, so the paper deserves peer review after a minor correction. read the letter →

arxiv 2507.14122 v1 pith:R2H5UL7O submitted 2025-07-18 math.OC

classification math.OC MSC 90C2590C15
keywords stochasticgradientdescentlast-iterateconvergenceconvexsmoothoptimizationvariance-freeanalysisinterpolationconstantiteratecomplexityminimaxrates
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Most guarantees for stochastic gradient descent in convex, smooth problems track the average of the iterates, not the point practitioners actually use, the last iterate. This paper closes that gap for plain SGD: under convexity, L-smoothness, and the single mild condition that the stochastic gradient has finite expected squared norm at some minimizer, the last-iterate expected function gap is O(ln T/√T) with the standard step size γ ≈ 1/(L√T). Earlier last-iterate bounds at this rate required uniform variance or bounded-gradient assumptions, which are hard to verify in practice. The result shows that no modification of SGD, no projection, and no uniform noise control is needed for a near-optimal last-iterate rate.

What carries the argument

The engine is a variance-transfer inequality: because the gradients ∇f_i are L-Lipschitz and the f_i are convex, E[‖∇f_i(x)‖²] ≤ 2L(1+ε)(f(x)-inf f) + (1+1/ε) σ_*² for every ε>0. This replaces a global uniform variance bound by the single constant σ_*² plus a function-gap term, which is exactly the quantity being bounded. The proof then feeds this inequality into a weighted telescoping lemma that converts pointwise descent inequalities into a last-iterate bound: it chooses convex weights α_t, telescopes the squared-distance terms in the SGD recursion, and isolates f(x_T)-inf f. The parameter ε is tuned as (1-γL)/(1+γL), making the coefficient of the function gap positive exactly when γL<1.

What would settle it

Run SGD with γ = 1/(2L√T) on any convex, L-smooth finite-sum or expectation problem satisfying σ_*² < ∞ and measure E[f(x_T)-inf f] over many seeds; if for some such problem the decay is slower than C ln(T)/√T for all constants C, the bound is false. A more targeted falsifier is an explicit, convex, L-smooth, non-interpolating example where σ_*² is finite but the last-iterate gap provably violates the stated upper bound, or a performance-estimation computation of the worst-case last-iterate gap exceeding the theorem's constant.

Watch

Extended reading notes

Core claim

The paper's central claim is that for a convex and L-smooth expected-risk problem with nonempty minimizer set, SGD's last iterate converges at the same near-optimal rate as its averaged iterates. If x* is a minimizer and σ_*² = E[‖∇f_i(x*)‖²] < ∞, then for any fixed horizon T ≥ 3 and step size with γL ∈ (0,1), E[f(x_T)-inf f] ≤ T^φ [2D²/(γ(1-γL)T) + 8γ ln(T+1) σ_*²/(1-γL)²], with φ = 2γL/(1+γL) and D² = E[‖x0-x*‖²]. Choosing γ proportional to 1/(L√T) makes T^φ a constant and yields E[f(x_T)-inf f] = O(ln T/√T). The only noise information needed is the variance at a minimizer; no uniform bound on the variance or norm of the stochastic gradients over the whole space is assumed. The paper presents these as the first such bounds while acknowledging a concurrent independent result with similar conclusions.

Load-bearing premise

The whole proof is carried by the assumption that σ_*², the expected squared norm of the stochastic gradient at some minimizer, is finite — if that single-point variance is infinite, the stated bound is vacuous — and by the step-size restriction γL < 1, so the endpoint γL = 1, which a concurrent work covers, is outside the theorem.

Editorial extensions

If this is right

  • With step size γ = 1/(C L √T), C ≥ 2, the explicit corollary gives E[f(x_T)-inf f] ≤ 9 C L D²/√T + 67 ln(T+1) σ_*²/(C L √T), matching the classical ergodic SGD rate up to a logarithm.
  • For any polynomial step size γ = 1/(C L T^β), β ∈ (0,1), the bound is O(D²/T^{1-β} + ln T σ_*²/T^β); β = 1/2 is the best exponent in this family.
  • The same framework extends to mini-batch SGD, because minibatch sampling is SGD on a modified problem whose smoothness constant and σ_*² are explicit functions of batch size.
  • A consequence is an iterate-complexity guarantee: E[f(x_T)-inf f] ≤ ε once T is of order (K/ε)^β for any β > 2, ignoring logarithms.
  • The bound is adaptive to the noise constant σ_*² in the sense that the algorithm need not know it to obtain the stated rate.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural next test is whether the logarithmic factor is removable: by analogy with the convex Lipschitz case, a nonstandard step-size schedule might give the exact O(1/√T) last-iterate rate; the paper leaves this as a conjecture.
  • The bound is not interpolation-adaptive: when σ_*² = 0 it remains O(1/√T) rather than O(1/T), so an open extension is a step-size rule that automatically switches to the fast rate when interpolation holds, using only quantities the algorithm can estimate.
  • It would be worth testing whether existing adaptive step-size rules inherit last-iterate guarantees under the same single-point variance condition; the paper notes this as an open direction.
  • A high-probability version of the last-iterate guarantee without uniform noise bounds does not follow from the expectation analysis and is not claimed by the paper; that is a concrete open problem.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. This paper studies constant-step-size SGD for minimizing f(x)=E[f_i(x)] where each f_i is convex and L-smooth. The main result (Theorem 3.1) states that, under Assumptions 2.1 and 2.2, E[f(x_T)-inf f] ≤ T^φ [2D^2/(γ(1-γL)T) + 8γ ln(T+1)/(1-γL)^2 σ_*^2], where φ=2γL/(1+γL) and γL∈(0,1); with γ≍1/√T this yields O(ln T/√T) last-iterate complexity. The proof combines a variance-transfer inequality at a minimizer (Lemma 4.1), a one-step recursion (Lemma 4.2), and a weighted telescoping argument (Lemma 4.3) adapted from Zamani-Glineur and Liu-Zhou. The paper also derives corollaries for polynomial step sizes and sample complexity, and acknowledges the concurrent work of Attia et al.

Significance. If the proof is made formally correct, this is a significant result: it removes the uniform gradient/variance assumption that previous last-iterate SGD bounds relied on and reduces the noise condition to finiteness of E||∇f_i(x*)||^2 at a minimizer. The paper proves its key technical lemmas in the text, gives explicit constants, and honestly describes the relation to parallel work; self-citations are used only for standard background inequalities. The O(ln T/√T) rate matches known ergodic rates up to a logarithm, and the authors correctly flag whether the logarithm is removable as open.

major comments (2)
  1. [Lemma 4.3 (Section 4.2)] The stated hypothesis -a < b ≤ 0 is incompatible with the application in Theorem 3.1. There a=(1-γL)/(1+γL)∈(0,1) and b=-1, so -a∈(-1,0) and the inequality -a < b becomes -a < -1, equivalently a>1, which is false for every γL∈(0,1). The proof's own construction p_t = α_{t-1}/α_t = (T-t+1+a/b)/(T-t+1) requires p_t∈[0,1]; with b<0 and a>0 this holds when 2+a/b>0, i.e. for b=-1 when 0<a<2. Thus the printed hypothesis describes the wrong side of the boundary and Lemma 4.3 cannot be invoked as stated. The algebra in the proof goes through under the corrected hypothesis b<0<a and 2+a/b>0; the authors should restate the lemma accordingly and re-verify the recurrence for α_t in that regime.
  2. [Lemma 4.3 and Lemma A.3 (Sections 4.2 and A)] The index ranges in the telescoping argument need to be harmonized. Lemma 4.3 defines α_t for t=1,...,T-1 via the recurrence, while the proof later sets α_T=α_{T-1} and sums the weighted inequality from 0 to T; Lemma A.3, invoked in Theorem 3.1, defines α_t for t=0,...,T with a recurrence that includes t=0. The identity α_T/α_{T-1}=1, which is used to obtain 4T^φ ln(T+1) from Lemma A.1, is not stated as part of the lemmas. Please specify the recurrence for α_0 and α_T and the exact summation range used to derive the final bound in Lemma 4.3.
minor comments (5)
  1. [Lemma 4.3 (Section 4.2)] The phrase 'for all zt ∈ H contained in F (x0, . . . , xt)' should be replaced by a clear measurability condition, e.g. 'for every z_t that is F(x0,...,xt)-measurable'.
  2. [Corollary 3.5] The sentence 'we take γ = 1/(2L√T), for some , and T (1+ln(T+1))^2 ≥ K^2/ε^2' contains a missing placeholder; it should read 'for some T' or similar.
  3. [Lemma A.3] The second displayed inequality is written with a parameter θ that appears to be a typo for φ; additionally, the definition 'for t=0,...,T' conflicts with the recurrence range in Lemma 4.3.
  4. [Theorem 3.1] The constant D^2 is not explicitly defined as an infimum over minimizers; since σ_*^2 is defined via an infimum, the bound should clarify that it holds for any x*∈argmin f with D^2=E||x0-x*||^2, or that D^2 is taken as the infimum over minimizers.
  5. [Proof of Lemma 4.2] The statement that z_t is 'independent from x_t' is imprecise; what is used is that z_t is measurable with respect to F(x0,...,xt), so that the conditional expectation E_t[⟨∇f_i(x_t), z_t⟩] = ⟨E_t[∇f_i(x_t)], z_t⟩ is valid.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central bound is derived in-text from variance-transfer and weighting lemmas; self-citations are background only. The printed Lemma 4.3 has a hypothesis mismatch with Theorem 3.1, but that is a proof-gap, not circularity.

full rationale

The claimed last-iterate rate is not an input to any definition or fit. Lemma 4.1 (variance transfer) is proved in the text from a Fenchel-Young inequality plus the standard cocoercivity inequality; the citation to (Garrigos and Gower, 2024) is to a handbook sharing an author, but the lemma is restated with a self-contained proof sketch, so the self-citation is not load-bearing. Lemma 4.2 is pure algebra from Lemma 4.1. Lemma 4.3 is the Zamani-Glineur weighting technique, reproduced with a full proof in the text rather than imported by citation; its use of Jensen with b <= 0 and telescoping does not assume the theorem's conclusion. Lemma A.3 (Gamma-function bounds) is proved via Gautschi's inequality, with the sum-integral comparison standard. No parameter is fitted to data and sigma_*^2 is a problem constant, not a fitted quantity renamed as a prediction. The concurrent-work note (Attia et al., 2025) only affects novelty, not circularity. The one notable defect is formal rather than circular: Lemma 4.3 states the hypothesis -a < b <= 0, while Theorem 3.1 applies it with b = -1 and a = (1 - gamma L)/(1 + gamma L) in (0,1), for which -a < b is false; the subsequent alpha_t recurrence is consistent with the actual regime (a/b in (-1,0)), so the central claim appears correct but the printed lemma needs its hypothesis corrected before it can be invoked. Because no derivation step reduces to its own input by construction, the circularity score is 0.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The paper introduces no fitted constants or new entities. The only hand-chosen scalar is the step-size gamma, an algorithmic input. The analysis relies on standard domain assumptions (convex, smooth, finite variance at minimizer) and standard mathematical tools.

free parameters (1)
  • step-size gamma = gamma = 1/(C L T^beta), beta in (0,1), C >= 2
    The step-size is a hyperparameter chosen by hand to balance the bias and variance terms in the rate; it is not inferred from data. The optimal exponent beta = 1/2 gives the headline O(log T / sqrt(T)) rate.
assumptions (4)
  • domain assumption Assumption 2.1: each f_i is convex and L-smooth, argmin f is nonempty, and sampling is well-defined.
    The variance-transfer Lemma 4.1 and the per-step inequalities in Lemma 4.2 both require convexity and L-smoothness of every component f_i.
  • domain assumption Assumption 2.2 (GV*): E[||grad f_i(x*)||^2] < infinity for some (equivalently every) minimizer x* of f, with sigma_*^2 = inf over minimizers.
    sigma_*^2 is the only noise constant in the final bound; without it the rate is vacuous. The paper argues it is mild (automatic for finite support and for nonnegative losses).
  • standard math Standard convex analysis tools: descent lemma for L-smooth functions, Jensen's inequality, Fenchel-Young, and Gautschi's inequality for the Gamma function.
    Used in Lemmas 4.1, 4.2, 4.3 and A.3; these are standard results, with Gautschi cited to Gautschi (1959).
  • domain assumption Step-size restriction gamma L in (0,1) for the generic theorem.
    The proof of Theorem 3.1 needs a = (1-gamma L)/(1+gamma L) > 0, which requires gamma L < 1; the boundary gamma L = 1 is excluded, though the concurrent Attia et al. result handles it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Last-Iterate Complexity of SGD for Convex and Smooth Stochastic Problems." pith.science (2026). https://pith.science/paper/R2H5UL7O

@misc{pith2026250714122,
  author       = {Pith},
  title        = {Pith review of: Last-Iterate Complexity of SGD for Convex and Smooth Stochastic Problems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R2H5UL7O}},
  note         = {Machine review of arXiv:2507.14122}
}
abstract

Most results on Stochastic Gradient Descent (SGD) in the convex and smooth setting are presented under the form of bounds on the ergodic function value gap. It is an open question whether bounds can be derived directly on the last iterate of SGD in this context. Recent advances suggest that it should be possible. For instance, it can be achieved by making the additional, yet unverifiable, assumption that the variance of the stochastic gradients is uniformly bounded. In this paper, we show that there is no need of such an assumption, and that SGD enjoys a $\tilde O \left( T^{-1/2} \right)$ last-iterate complexity rate for convex smooth stochastic problems.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Fast Last-Iterate Convergence of SGD in the Smooth Interpolation Regime

    cs.LG 2025-07 accept novelty 7.0 of 10

    The last SGD iterate converges at near-optimal rates in the smooth convex interpolation regime, including O(1/sqrt(T)) with the greedy step size eta=1/beta for general convex smooth losses.

Reference graph

Works this paper leans on

30 extracted references · 22 canonical work pages · cited by 1 Pith paper

  1. [1]

    Alacaoglu, A., Malitsky, Y., and Wright, S. J. (2025). Towards Weaker Variance Assumptions for Stochastic Optimization . arXiv preprint arXiv:2504.09951

  2. [2]

    Attia, A., Schliserman, M., Sherman, U., and Koren, T. (2025). Fast Last-Iterate Convergence of SGD in the Smooth Interpolation Regime . arXiv preprint arXiv:2507.11274

  3. [3]

    and Moulines, E

    Bach, F. and Moulines, E. (2011). Non- Asymptotic Analysis of Stochastic Approximation Algorithms for Machine Learning . In Advances in Neural Information Processing Systems , volume 24. Curran Associates, Inc

  4. [4]

    Blum, J. R. (1954). Approximation Methods which Converge with Probability one. The Annals of Mathematical Statistics , 25(2):382--386

  5. [5]

    E., and Nocedal, J

    Bottou, L., Curtis, F. E., and Nocedal, J. (2018). Optimization Methods for Large - Scale Machine Learning . SIAM Review , 60(2):223--311

  6. [6]

    Cortild, D., Ketels, L., Peypouquet, J., and Garrigos, G. (2025). New Tight Bounds for SGD without Variance Assumption : A Computer - Aided Lyapunov Analysis . arXiv preprint arXiv:2505.17965

  7. [7]

    and Gower, R

    Garrigos, G. and Gower, R. M. (2024). Handbook of Convergence Theorems for ( Stochastic ) Gradient Methods . arXiv preprint arXiv:2301.11235

  8. [8]

    Gautschi, W. (1959). Some Elementary Inequalities Relating to the Gamma and Incomplete Gamma Function . Journal of Mathematics and Physics , 38(1-4):77--81

Show all 30 references
  1. [9]

    Gladyshev, E. G. (1965). On Stochastic Approximation . Theory of Probability & Its Applications , 10(2):275--278

  2. [10]

    Gower, R., Sebbouh, O., and Loizou, N. (2021). SGD for Structured Nonconvex Functions : Learning Rates , Minibatching and Interpolation . In Proceedings of the 24th International Conference on Artificial Intelligence and Statistics , pages 1315--1323. PMLR

  3. [11]

    M., Garrigos, G., Loizou, N., Oikonomou, D., Mishchenko, K., and Schaipp, F

    Gower, R. M., Garrigos, G., Loizou, N., Oikonomou, D., Mishchenko, K., and Schaipp, F. (2025). Analysis of an idealized stochastic Polyak method and its application to black-box model distillation . arXiv preprint arXiv:2504.01898

  4. [12]

    M., Loizou, N., Qian, X., Sailanbayev, A., Shulgin, E., and Richtárik, P

    Gower, R. M., Loizou, N., Qian, X., Sailanbayev, A., Shulgin, E., and Richtárik, P. (2019). SGD : General Analysis and Improved Rates . In Proceedings of the 36th International Conference on Machine Learning , pages 5200--5209. PMLR

  5. [13]

    Harvey, N. J. A., Liaw, C., Plan, Y., and Randhawa, S. (2018). Tight Analyses for Non - Smooth Stochastic Gradient Descent . arXiv preprint arXiv:1812.05217

  6. [14]

    Jain, P., Nagaraj, D., and Netrapalli, P. (2019). Making the Last Iterate of SGD Information Theoretically Optimal . arXiv preprint arXiv:1904.12443

  7. [15]

    Kassing, S., Weissmann, S., and D \"o ring, L. (2025). Controlling the Flow: Stability and Convergence for Stochastic Gradient Descent with Decaying Regularization . arXiv preprint arXiv:2505.11434

  8. [16]

    and Richtárik, P

    Khaled, A. and Richtárik, P. (2023). Better Theory for SGD in the Nonconvex World . Transactions on Machine Learning Research

  9. [17]

    D., Nguyen, T

    Liu, Z., Nguyen, T. D., Nguyen, T. H., Ene, A., and Nguyen, H. (2023). High Probability Convergence of Stochastic Gradient Methods . In Proceedings of the 40th International Conference on Machine Learning . PMLR

  10. [18]

    and Zhou, Z

    Liu, Z. and Zhou, Z. (2023). Revisiting the Last - Iterate Convergence of Stochastic Gradient Methods . In Proceedings of The Twelfth International Conference on Learning Representations

  11. [19]

    H., and Lacoste-Julien, S

    Loizou, N., Vaswani, S., Laradji, I. H., and Lacoste-Julien, S. (2021). Stochastic Polyak Step-size for SGD: An Adaptive Learning Rate for Fast Convergence . In International Conference on Artificial Intelligence and Statistics , pages 1306--1314. PMLR

  12. [20]

    How do you prove Gautschi's inequality for the gamma function? https://math.stackexchange.com/q/98348

    Mathematics Stack Exchange (2017). How do you prove Gautschi's inequality for the gamma function? https://math.stackexchange.com/q/98348

  13. [21]

    Needell, D., Srebro, N., and Ward, R. (2016). Stochastic gradient descent, weighted sampling, and the randomized Kaczmarz algorithm. Mathematical Programming , 155(1):549--573

  14. [22]

    Nemirovski, A., Juditsky, A., Lan, G., and Shapiro, A. (2009). Robust Stochastic Approximation Approach to Stochastic Programming . SIAM Journal on Optimization , 19(4):1574--1609

  15. [23]

    H., Dijk, M., Richtarik, P., Scheinberg, K., and Takac, M

    Nguyen, L., Nguyen, P. H., Dijk, M., Richtarik, P., Scheinberg, K., and Takac, M. (2018). SGD and Hogwild ! Convergence Without the Bounded Gradients Assumption . In Proceedings of the 35th International Conference on Machine Learning , pages 3750--3758. PMLR

  16. [24]

    Orabona, F. (2020). Last Iterate of SGD Converges (Even in Unbounded Domains) . https://parameterfree.com/2020/08/07/last-iterate-of-sgd-converges-even-in-unbounded-domains/

  17. [25]

    and D'Orazio, R

    Orabona, F. and D'Orazio, R. (2025). New Perspectives on the Polyak Stepsize: Surrogate Functions and Negative Results . arXiv preprint arXiv:2505.20219

  18. [26]

    and Monro, S

    Robbins, H. and Monro, S. (1951). A Stochastic Approximation Method . The Annals of Mathematical Statistics , 22(3):400--407

  19. [27]

    M., and Defazio, A

    Sebbouh, O., Gower, R. M., and Defazio, A. (2021). Almost sure convergence rates for stochastic gradient descent and stochastic heavy ball. In Conference on Learning Theory , pages 3935--3971. PMLR

  20. [28]

    and McMahan, H

    Streeter, M. and McMahan, H. B. (2010). Less regret via online conditioning. arXiv preprint arXiv:1002.4862

  21. [29]

    and Bach, F

    Taylor, A. and Bach, F. (2019). Stochastic first-order methods: non-asymptotic and computer-aided analyses via potential functions. In Proceedings of the 32nd Conference on Learning Theory , pages 2934--2992. PMLR

  22. [30]

    and Glineur, F

    Zamani, M. and Glineur, F. (2023). Exact convergence rate of the last iterate in subgradient methods. arXiv preprint arXiv:2307.11134

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.