Specular differentiation in normed vector spaces: Quasi-Mean Value and Quasi-Fermat Theorems

Kiyuob Jung

REVIEW 4 major objections 3 minor 3 references

Specular differentiation in normed vector spaces: Quasi-Mean Value and Quasi-Fermat Theorems

T0 review · 4 major / 3 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper introduces specular differentiation, a mirror-inspired generalization of Gâteaux and Fréchet derivatives in normed vector spaces, and shows that for convex functions its gradient is a subgradient, yielding new optimization method

desk verdict Specular gradient is a genuine subgradient for convex functions, but the stochastic convergence theory rests on a false unbiasedness condition and the numerical comparison omits the obvious baseline. read the letter →

arxiv 2601.10950 v3 pith:RM4RTHCA submitted 2026-01-16 math.OC

Kiyuob Jung This is my paper · ORCID

classification math.OC MSC 46G0546T2049J5265K0590C2590C15

keywords speculardifferentiationgeneralizedderivativeQuasi-MeanValueTheoremQuasi-Fermatnonsmoothconvexoptimizationsubgradientmethodstochasticsubdifferential

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

The paper's central project is to define a new kind of derivative, the specular derivative, built from a weighted average of forward and backward difference quotients with weights chosen by distances on the graph of the function. It proves that whenever an ordinary Gâteaux or Fréchet derivative exists, the specular derivative exists and matches it, so the new notion is a true generalization rather than a competing rival. The key payoff is in nonsmooth convex optimization: in R^n, if a convex function is specularly Fréchet differentiable, its specular gradient is a subgradient, so the specular gradient method inherits the convergence guarantees of the classical subgradient method. The paper also establishes weak, inequality-based versions of the Mean Value Theorem and Fermat's Theorem in normed vector spaces, and demonstrates numerically that specular-gradient methods minimize elastic-net problems where gradient descent, Adam, and BFGS fail.

What carries the argument

The central object is the specular directional derivative and its equivalent half-angle formula in terms of one-sided derivatives. The formula ∂∧_v f(x) = ∥v∥ tan(½ arctan(∂+_v f(x)/∥v∥) + ½ arctan(∂−_v f(x)/∥v∥)) converts left and right derivative information into a single number that always lies between them, and degenerates to the classical derivative when both sides agree. When f is convex, this interpolation guarantees the specular gradient satisfies the subgradient inequality, which is the mechanism that imports all subgradient-method convergence theory. The auxiliary functions A and B encode the same identity in algebraic form for estimates and proofs.

What would settle it

Compute, at x = 0, the specular gradient of f(x) = (|x| + 2x)/2 directly from the half-angle formula (≈0.884) and compare it with the average of the specular gradients of |x| (which is 0) and 2x (which is 2), each sampled with probability 1/2. The two numbers differ (1 vs ≈0.884), so the unbiasedness condition (3.13) is violated, and any convergence theorem for the stochastic method that relies on it does not follow.

Watch

Extended reading notes

Core claim

The discovery at the center of the paper is the specular directional derivative ∂∧_v f(x), formed by reflecting the forward and backward secant lines about a mirror in the graph of f; in R^n it equals ∥v∥ tan(½ arctan(∂+_v f(x)/∥v∥) + ½ arctan(∂−_v f(x)/∥v∥)). When the classical directional derivative exists, the specular derivative agrees with it, and in normed vector spaces specular Gâteaux and Fréchet differentiability generalize the classical notions. The load-bearing result is Theorem 3.2: for a convex function on R^n that is specularly Fréchet differentiable, the specular gradient ∇s f(x) is a genuine subgradient, i.e., f(w) ≥ f(x) + ∇s f(x)·(w−x) for all w. Combined with the Quasi-Fer

Load-bearing premise

The stochastic specular gradient method rests on the unproven assumption that averaging specular gradients of randomly chosen component functions gives the specular gradient of the whole function, an equality that is not proven and is false for simple convex decompositions.

Editorial extensions

If this is right

For any convex, specularly Fréchet differentiable function on R^n, the specular gradient gives a deterministic rule for choosing a subgradient, so subgradient-method convergence theorems apply without separately computing the subdifferential.
In normed vector spaces, local extrema of specularly Gâteaux differentiable functionals satisfy the explicit necessary condition |∂∧_v f(x*)| ≤ ∥v∥, a quantitative generalization of Fermat's theorem to nondifferentiable functions.
The Quasi-Mean Value Theorem bounds the change in f between two points by the supremum and infimum of specular directional derivatives along the segment, giving a substitute for the classical mean value theorem when f is not differentiable.
The specular gradient method with square-summable, non-summable step sizes converges to a minimizer, matching subgradient-method worst-case guarantees and, per the numerical experiments, outperforming gradient descent, Adam, and BFGS on elastic-net instances where those methods fail.
The hybrid and stochastic variants improve wall-clock time while retaining convergence, provided the stochastic unbiasedness assumption holds.

Reading between the lines

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

A caution that goes beyond the paper's claims: the unbiasedness condition E[∇s f_ξ(x)|x_k] = ∇s f(x_k) assumed in Definition 3.7 is not established in the paper, and because the specular derivative is a nonlinear function of function values it can fail for natural decompositions; for f = (|x| + 2x)/2 at x = 0, sampling f1 = |x| and f2 = 2x with equal probability gives E[∇s f_ξ] = 1 while ∇s f ≈ 0.
A testable design consequence: the half-angle formula suggests specular gradients are a one-parameter family of secant-like interpolants; one could generalize by replacing the arithmetic mean of arctangents with other symmetric means, producing a continuum of specular derivatives with adjustable bias between left and right slopes.
Because Theorem 2.12 requires a norm-preserving linear functional on Y, an assumption that fails for many infinite-dimensional spaces, the Quasi-Mean Value Theorem is most naturally stated for real-valued or Hilbert-space-valued maps; extending it to general Banach targets would require a different mechanism.
The equivalence in Corollary 2.8 between existence of one-sided derivatives and specular differentiability suggests a fast differentiability check: a function is specularly differentiable exactly when its left and right slopes exist and are not both +∞ or both −∞, which may ease automatic differentiation of nonsmooth programs.

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.

Desk Editor's Note

read the letter

Colleague, here's my read of arXiv:2601.10950. The genuinely new and useful core is small but real: for convex functions on R^n, if f is specularly Fréchet differentiable at x then the specular gradient ∇s f(x) is a subgradient (Thm 3.2), and the same idea works in Banach spaces via the Fréchet subdifferential (Thm 2.19). The chain of reasoning from the weighted-slope definition through the A-function representation to the subgradient inclusion is coherent; I don't see a hole in that part. The paper also ships code and is candid that the pure specular method is slower in wall-clock terms than classical methods, which is more honest than most optimization papers.

The soft spots are real. The stochastic section does not hold together. Definition 3.7/eq. (3.13) assumes E[∇s f_ξ(x_k)|x_k] = ∇s f(x_k), but specular differentiation is not additive, so for f = (1/m)Σf_j the conditional expectation is (1/m)Σ∇s f_j, not ∇s f. The paper gives no proof and the condition is false for simple component sums; the stress-test example with f1=|x|, f2=2x at 0 works. Theorem 3.12's convergence in expectation and probability therefore has no valid proof. The deterministic SPEG convergence (Thm 3.10) is fine, but the advertised S-SPEG and H-SPEG results are not backed by the theory.

Also, Corollary 3.3(a) claims equality ⟺ x=x*; for a convex function constant on an interval, every interior point has zero specular gradient, so equality holds for distinct minimizers. The equality needs strict convexity or an extra argument. And Theorem 2.12's assumption (2.20) essentially forces Y to be one-dimensional, so the normed-space Quasi-MVT reduces to the R case; the generality is decorative rather than substantive.

The numerical section never compares against the plain subgradient method, which is the natural baseline since SPEG is just a subgradient selection. Showing that GD and Adam fail on a highly nonsmooth elastic-net instance is not the same as showing an advantage over a standard nonsmooth solver.

Bottom line: the subgradient-inclusion result and its deterministic method deserve to be known, but the stochastic claims in their current form are not reliable. A referee should ask for a corrected stochastic analysis or a weakened statement, a fix to Cor 3.3(a), and a comparison against SG. That is a substantial revision, not a desk reject.

Referee Report

4 major / 3 minor

Summary. The paper introduces "specular differentiation" in normed vector spaces, a nonlinear weighted-average derivative that generalizes Gâteaux and Fréchet differentiability when the latter exist. It proves several structural results: a formula for specular directional derivatives in terms of one-sided derivatives, a Quasi-Mean Value Theorem and Quasi-Fermat Theorem, inclusion of the specular Fréchet differential in the convex subdifferential, and convergence of deterministic and stochastic specular-gradient methods for nonsmooth convex optimization in R^n. The numerical section compares SPEG, S-SPEG, and H-SPEG against GD, Adam, and BFGS on an elastic-net objective, reporting that the proposed methods succeed where classical methods sometimes fail.

Significance. If sound, the deterministic part of the paper would be a useful contribution: it gives an explicit, computable construction of a subgradient for convex functions, a clean arctangent representation, and a public Python package. The convergence proof for the deterministic specular-gradient method is a straightforward but valid adaptation of subgradient-method arguments. However, several load-bearing claims are not correct as stated: the characterization theorem for specular directional derivatives is false in general, the Quasi-Mean Value Theorem is vacuous for codomains of dimension greater than one, and the stochastic convergence theorem does not apply to the implemented component-sampling algorithm. These issues are substantial and require more than cosmetic revision.

major comments (4)

[§2.1, Theorem 2.5/2.7] The converse direction of Theorem 2.5 is false. Take f(0)=0 and f(x)=|x|(1+sin(1/|x|)) for x≠0. The two weighted terms in (1.2) cancel for every h, so ∂∧_1 f(0)=0 and in fact D_s f(0)=0 (specular Fréchet differentiability holds), but the one-sided derivatives do not exist. Thus (H1) is not necessary for ∂∧_v f(x) to exist. Theorem 2.7, which assumes only existence of ∂∧_v f(x) and then uses ∂±_v f(x), is therefore false as stated. The proof uses a bound |...-c|<h that is not implied by the limit. The forward implication (H1)+(H2)⇒(2.5) appears correct; please restate the theorem as a sufficient condition or add hypotheses that guarantee one-sided differentiability.
[§2.2, Theorem 2.12] Assumption (2.20) forces Y to be one-dimensional: if φ(y)=0 then ||y||=|φ(y)|=0, so φ is injective from Y to R. Hence the Quasi-Mean Value Theorem does not generalize to vector-valued maps in any nontrivial Banach codomain; it is essentially the one-dimensional result from [5] applied to g=⟨φ,f⟩. The proof also assumes ⟨φ,f(v)-f(u)⟩ = ||f(v)-f(u)||, but a φ satisfying (2.20) may give the opposite sign. Please state the theorem for Y=R (or one-dimensional Y) and explicitly choose the sign of φ.
[§3.2, Definition 3.7/Theorem 3.12] Condition (3.13) is an unverified and generally false premise for the implemented algorithm. For f=(1/m)Σ f_j and uniform sampling, E[∇s f_ξ(x)]=(1/m)Σ ∇s f_j(x), but the specular derivative is nonlinear in the component slopes (Theorem 2.5), so this average need not equal ∇s f(x). Concrete failure at x=0: f1=|x|+x², f2=2x+x². Then ∇s f1(0)=0, ∇s f2(0)=2, so E=1, while ∇s f(0)=tan(½ arctan(3/2)+½ arctan(1/2))≈0.884. Thus the replacement of E[∇s f_ξ(x_ℓ)|x_ℓ] by ∇s f(x_ℓ) in the proof of Theorem 3.12 is unjustified for the S-SPEG/H-SPEG experiments in §3.2. Additionally, part (a) is false: a zero component specular gradient does not imply a minimizer of f; e.g., f1=(x)_+², f2=(x-1)² at x=0.
[§3.1, Corollary 3.3(a)] The assertion that equality in ∇s f(x)·(x−x*)≥0 holds if and only if x=x* is false. For f(x)=max(x,0)^2 on Ω=R and x*=0, every x<0 satisfies ∇s f(x)=0, so the dot product is zero even though x≠x*. The proof only shows that equality implies f(x)=f(x*), not x=x*. The statement should be weakened accordingly.

minor comments (3)

[Algorithm 3.1 and Algorithm 3.2] The break condition appears reversed: the pseudocode stops when ∥∇s f(x)∥ > η, but it should stop when the gradient norm is below tolerance (≤ η). As printed, the algorithms would terminate immediately at most starting points.
[§3.2] The notation x⋄_k is defined in (3.8) for the deterministic SPEG method only, but Theorem 3.12 uses x⋄_k for S-SPEG without restating the definition. Define it for the stochastic iterates.
[§2.1, Remark 2.4] The statement that the one-sided limit (h↘0) in (2.4) exists if and only if the limit (h→0) exists is ambiguous, since the definition only uses h>0. Please clarify the intended two-sided statement or remove the remark.

Circularity Check

1 steps flagged · score 2.0 of 10

No significant circularity in the main derivation chain; one tautological proposition and an unproven stochastic unbiasedness condition are the only concerns.

self definitional [Section 3.1, Proposition 3.4]
"Proposition 3.4. Let Ω be an open convex set in R^n. If a function f: Ω→R is convex and specularly Fréchet differentiable in Ω and satisfies that (3.4) f(w) ≥ ∇s f(x)·(w−x)+f(x) for all x,w∈Ω, then f is convex in Ω."

The conclusion 'f is convex in Ω' is already stated as a hypothesis, making the proposition a tautology as written. The proof actually derives convexity from inequality (3.4) alone, so the convexity hypothesis is redundant; the statement asserts what it assumes and contributes no independent content.

full rationale

The core derivation is otherwise self-contained: Propositions 2.10 and 2.15 derive compatibility with Gâteaux and Fréchet differentiability from Definition 1.1; Theorem 2.19 and Theorem 3.2 prove the subgradient inclusion from convexity and the specular formula; and Theorem 3.10 is a standard subgradient-method convergence argument using Theorem 3.2. The self-citations to [5] provide parameter-free one-dimensional lemmas and the 1D Quasi-MVT; these are load-bearing but independent, falsifiable results, so they do not constitute circularity. The stochastic theorem (Theorem 3.12) is conditional on the unbiasedness assumption (3.13), which is unproven and fails for the natural uniform component sampler because the specular derivative is nonlinear; this is a substantive correctness concern, not a circular reduction. The score of 2 reflects only the minor tautological Proposition 3.4.

Assumptions & free parameters 5 free parameters · 5 assumptions · 2 invented entities

The theory itself has no fitted physical constants; the free parameters listed are experimental choices (step sizes, switching iteration) that influence the numerical claims. The main load-bearing axioms are convexity plus specular Fréchet differentiability, the restrictive norming assumption (2.20) in the MVT, and the unverified unbiasedness condition (3.13) in the stochastic method.

free parameters (5)

SPEG-s step-size constant = 4
Step size h_k = 4 / ((k+1)||∇s f(x_k)||) chosen by hand; affects convergence speed in experiments.
SPEG-g step-size ratio = 1/2
Geometric step size h_k = 2^{-k-1}; experiment shows it diverges in R^n.
GD learning rate = 0.001
Chosen empirically for comparison baseline.
Adam learning rate = 0.01
Chosen empirically for comparison baseline.
H-SPEG switching iteration = 10
Switches from SPEG to S-SPEG after k=10 iterations; chosen without sensitivity analysis.

assumptions (5)

domain assumption f is convex and specularly Fréchet differentiable in Ω
Assumed in Lemmas 3.1/3.9 and Theorems 3.2/3.10/3.12; the paper explicitly leaves open whether the assumption can be relaxed.
ad hoc to paper Existence of φ∈Y* with ||φ||=1 and |⟨φ,y⟩|=||y|| for all y∈Y (assumption (2.20))
Used in Theorem 2.12 (Quasi-MVT). This condition forces Y to be one-dimensional, making the normed-space generalization vacuous.
ad hoc to paper Unbiasedness condition E[∇s f_ξ(x)|x_k] = ∇s f(x_k) (eq. 3.13)
Assumed without proof in Definition 3.7; generally false for nonlinear specular derivatives of component sums, as shown by the |x| and 2x example.
domain assumption Bounded specular gradients (H5): ||∇s f_j(x_k)|| ≤ C
Assumed in Theorem 3.12 to control the stochastic recursion.
standard math Standard functional analysis results: Hahn-Banach, Riesz representation, Bolzano-Weierstrass, Rockafellar's convex subdifferential characterizations
Used implicitly or via references throughout Sections 2 and 3.

invented entities (2)

Specular derivative / specular directional derivative
purpose: Generalized derivative for nonsmooth functions, defined by a mirror-weighted average of left and right difference quotients (Definition 1.1).
A mathematical construction introduced by the authors; it has proven consistency properties inside the paper but no external falsifiable handle.
Specular gradient
purpose: A selected subgradient for convex functions, obtained from the specular Fréchet differential via the Riesz representation (Definition 1.3).
Derived mathematical object; its usefulness depends on the subgradient theorem, which is proven in the paper but lacks independent verification outside the authors' framework.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Specular differentiation in normed vector spaces: Quasi-Mean Value and Quasi-Fermat Theorems." pith.science (2026). https://pith.science/paper/RM4RTHCA

@misc{pith2026260110950,
  author       = {Pith},
  title        = {Pith review of: Specular differentiation in normed vector spaces: Quasi-Mean Value and Quasi-Fermat Theorems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RM4RTHCA}},
  note         = {Machine review of arXiv:2601.10950}
}

read the original abstract

This paper introduces specular differentiation, which generalizes G\^ateaux and Fr\'echet differentiation in normed vector spaces. We investigate its fundamental theoretical properties and establish weak forms of the Mean Value Theorem and Fermat's Theorem in the specular sense. Finally, we identify a distinguished element of the Fr\'echet subdifferential of a convex function through specular differentiation.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

3 extracted references · 2 linked inside Pith

[1]

Boyd,Subgradient methods, lecture notes of EE364b, Stanford University, (2014), https: //web.stanford.edu/class/ee364b/lectures/subgrad method notes.pdf

[1]S. Boyd,Subgradient methods, lecture notes of EE364b, Stanford University, (2014), https: //web.stanford.edu/class/ee364b/lectures/subgrad method notes.pdf. [2]S. Boyd, A. Mutapcic, and J. Duchi,Stochastic subgradient methods, lecture notes of EE364b, Stanford University, (2018), https://web.stanford.edu/class/ee364b/lectures/ stoch subgrad notes.pdf. ...

arXiv 2014
[1970]

[16]R. T. Rockafellar and R. J.-B. Wets,Variational analysis, vol. 317, Springer-Verlag, Berlin, 1998, https://doi.org/10.1007/978-3-642-02431-3. [17]A. Ruszczy ´nski,Nonlinear optimization, Princeton University Press, Princeton, NJ,

doi:10.1007/978-3-642-02431-3 1998
[2023]

Schirotzek,Nonsmooth analysis, Universitext, Springer, Berlin, 2007, https://doi.org/10

[19]W. Schirotzek,Nonsmooth analysis, Universitext, Springer, Berlin, 2007, https://doi.org/10. 1007/978-3-540-71333-3. [20]N. Z. Shor,Minimization methods for nondifferentiable functions, vol. 3 of Springer Series in Computational Mathematics, Springer-Verlag, Berlin, 1985, https://doi.org/10.1007/ 978-3-642-82118-9. [21]N. Z. Shor,Nondifferentiable opti...

arXiv 2007

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

Boyd,Subgradient methods, lecture notes of EE364b, Stanford University, (2014), https: //web.stanford.edu/class/ee364b/lectures/subgrad method notes.pdf

[1]S. Boyd,Subgradient methods, lecture notes of EE364b, Stanford University, (2014), https: //web.stanford.edu/class/ee364b/lectures/subgrad method notes.pdf. [2]S. Boyd, A. Mutapcic, and J. Duchi,Stochastic subgradient methods, lecture notes of EE364b, Stanford University, (2018), https://web.stanford.edu/class/ee364b/lectures/ stoch subgrad notes.pdf. ...

arXiv 2014

[2] [1970]

[16]R. T. Rockafellar and R. J.-B. Wets,Variational analysis, vol. 317, Springer-Verlag, Berlin, 1998, https://doi.org/10.1007/978-3-642-02431-3. [17]A. Ruszczy ´nski,Nonlinear optimization, Princeton University Press, Princeton, NJ,

doi:10.1007/978-3-642-02431-3 1998

[3] [2023]

Schirotzek,Nonsmooth analysis, Universitext, Springer, Berlin, 2007, https://doi.org/10

[19]W. Schirotzek,Nonsmooth analysis, Universitext, Springer, Berlin, 2007, https://doi.org/10. 1007/978-3-540-71333-3. [20]N. Z. Shor,Minimization methods for nondifferentiable functions, vol. 3 of Springer Series in Computational Mathematics, Springer-Verlag, Berlin, 1985, https://doi.org/10.1007/ 978-3-642-82118-9. [21]N. Z. Shor,Nondifferentiable opti...

arXiv 2007