Recognition: unknown
Almost-Orthogonality in Lp Spaces: A Case Study with Grok
Pith reviewed 2026-05-08 15:59 UTC · model grok-4.3
The pith
Carbery's sharpened triangle inequality in Lp fails for all p>2, but holds at the critical exponent c=p' for integer p>=2, with an optimal three-function version.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
The central claim is that the inequality ||sum f_j||_p ≤ (sup_j sum_k alpha_jk^c)^{1/p'} (sum ||f_j||_p^p)^{1/p} fails for c=2 and every p>2, that any valid version must obey c ≤ p', that the inequality holds for all integer p≥2 when c=p', and that the three-function case admits the sharp bound ||sum_{j=1}^3 f_j||_p ≤ (1 + 2 Gamma^{c(p)})^{1/p'} (sum ||f_j||_p^p)^{1/p} with c(p) = 2 ln(2)/((p-2)ln(3)+2 ln(2)) that cannot be improved.
What carries the argument
The pairwise almost-orthogonality coefficient alpha_jk = sqrt( ||f_j f_k||_{p/2} / (||f_j||_p ||f_k||_p) ) raised to a power c inside a supremum that controls the norm of the sum.
If this is right
- The exponent in any almost-orthogonality bound of this type cannot exceed p' or the inequality is false in general.
- For every integer p at least 2 the inequality holds with c equal to p'.
- For three functions the constant 1+2 Gamma^{c(p)} is sharp and strictly improves the earlier exponent 6/(5p-4).
- The counterexample rules out the original c=2 proposal for all p>2.
Where Pith is reading between the lines
- The three-function result with its explicit optimal c(p) supplies a template for deriving similar sharp constants when more than three functions are involved.
- Applications that bound sums of Lp functions under partial orthogonality, such as in harmonic analysis or PDE estimates, can now safely use c up to p' for integer p.
- The paper's use of an AI model to explore intermediate inequalities indicates a possible workflow for verifying or discovering exponents in related functional inequalities.
Load-bearing premise
The functions lie in Lp so that the normalized p/2-norm products defining alpha_jk are finite and the overlap quantity Gamma can be chosen in [0,1].
What would settle it
Explicit functions in L^3 whose pairwise products yield alpha_jk such that the left-hand side of the proposed inequality exceeds the right-hand side for c=2, or three functions attaining equality in the c(p) bound for p=3.
read the original abstract
Carbery proposed the following sharpened form of triangle inequality for many functions: for any $p\ge 2$ and any finite sequence $(f_j)_j\subset L^p$ we have \[ \Big\|\sum_j f_j\Big\|_p \ \le\ \left(\sup_{j} \sum_{k} \alpha_{jk}^{\,c}\right)^{1/p'} \Big(\sum_j \|f_j\|_p^p\Big)^{1/p}, \] where $c=2$, $1/p+1/p'=1$, and $\alpha_{jk}=\sqrt{\frac{\|f_{j}f_{k}\|_{p/2}}{\|f_{j}\|_{p}\|f_{k}\|_{p}}}$. In the first part of this paper we construct a counterexample showing that this inequality fails for every $p>2$. We then prove that if an estimate of the above form holds, the exponent must satisfy $c\le p'$. Finally, at the critical exponent $c=p'$, we establish the inequality for all integer values $p\ge 2$. In the second part of the paper we obtain a sharp three-function bound \[ \Big\|\sum_{j=1}^{3} f_j\Big\|_p \ \le\ \left(1+2\Gamma^{c(p)}\right)^{1/p'} \Big(\sum_{j=1}^{3} \|f_j\|_p^p\Big)^{1/p}, \] where $p \geq 3$, $c(p) = \frac{2\ln(2)}{(p-2)\ln(3)+2\ln(2)}$ and $\Gamma=\Gamma(f_1,f_2,f_3)\in[0,1]$ quantifies the degree of orthogonality among $f_1,f_2,f_3$. The exponent $c(p)$ is optimal, and improves upon the power $r(p) = \frac{6}{5p-4}$ obtained previously by Carlen, Frank, and Lieb. Some intermediate lemmas and inequalities appearing in this work were explored with the assistance of the large language model Grok.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript studies Carbery's proposed sharpened triangle inequality in L^p spaces (p ≥ 2) that incorporates an almost-orthogonality factor α_jk defined via the normalized L^{p/2} product norm. It constructs an explicit counterexample showing failure for the exponent c=2 when p>2, proves that any valid inequality of this form requires c ≤ p', establishes the inequality at the critical value c=p' for all integer p ≥ 2, and derives a sharp three-function bound with optimal exponent c(p) = 2 ln(2)/((p-2) ln(3) + 2 ln(2)) that improves on the Carlen-Frank-Lieb exponent r(p) = 6/(5p-4).
Significance. If the counterexample and proofs hold, the work supplies concrete, falsifiable information on the admissible range of exponents for almost-orthogonality inequalities, together with an optimal three-function constant derived from logarithmic optimization. The explicit constructions and the improvement over prior exponents constitute the main advance.
major comments (2)
- [Necessity proof] The necessity argument that c must satisfy c ≤ p' is load-bearing for the first part of the paper; the abstract states it follows from standard L^p properties, but the precise reduction (e.g., via a suitable choice of two or three functions that forces the exponent bound) should be written out explicitly to confirm it does not rely on additional assumptions about the support or normalization of the f_j.
- [Three-function estimate] For the three-function bound, the optimality claim for c(p) rests on a logarithmic optimization that produces the explicit formula; the manuscript should verify that this c(p) is indeed attained by some choice of f1,f2,f3 in L^p with Γ(f1,f2,f3) ∈ (0,1) and that the resulting constant is strictly smaller than the Carlen-Frank-Lieb value for p ≥ 3.
minor comments (2)
- [Abstract] The abstract mentions that some intermediate lemmas were explored with the assistance of Grok; if any of those lemmas are used in the published proofs, a brief statement of which steps were machine-assisted would improve transparency without altering the mathematical content.
- [Three-function bound] Notation for Γ(f1,f2,f3) is introduced as an orthogonality measure in [0,1]; a short paragraph recalling its precise definition (presumably via the α_jk quantities) would help readers who have not yet reached the three-function section.
Simulated Author's Rebuttal
We are grateful to the referee for the positive assessment and the detailed suggestions for improvement. We will revise the manuscript to address the two major comments as outlined below.
read point-by-point responses
-
Referee: [Necessity proof] The necessity argument that c must satisfy c ≤ p' is load-bearing for the first part of the paper; the abstract states it follows from standard L^p properties, but the precise reduction (e.g., via a suitable choice of two or three functions that forces the exponent bound) should be written out explicitly to confirm it does not rely on additional assumptions about the support or normalization of the f_j.
Authors: We agree that expanding the necessity argument will enhance clarity. In the revised manuscript, we will provide an explicit reduction by constructing suitable functions (for instance, two or three functions with disjoint or partially overlapping supports in a measure space) that force the exponent to satisfy c ≤ p'. This construction relies solely on the definition of the L^p norm, the almost-orthogonality factor α_jk, and Hölder's inequality, without additional assumptions on normalization beyond the standard unit-norm setting. We will insert this as a dedicated lemma or subsection following the statement of the necessity result. revision: yes
-
Referee: [Three-function estimate] For the three-function bound, the optimality claim for c(p) rests on a logarithmic optimization that produces the explicit formula; the manuscript should verify that this c(p) is indeed attained by some choice of f1,f2,f3 in L^p with Γ(f1,f2,f3) ∈ (0,1) and that the resulting constant is strictly smaller than the Carlen-Frank-Lieb value for p ≥ 3.
Authors: We will incorporate a verification of the optimality of c(p) in the revised version. We will present an explicit example of functions f1, f2, f3 in L^p (p ≥ 3) for which Γ(f1,f2,f3) lies strictly between 0 and 1, and demonstrate that equality holds in the three-function inequality precisely when the exponent is c(p), thereby confirming that the bound is sharp. Furthermore, we will add a short computation or table comparing c(p) and r(p) = 6/(5p-4) for several values of p ≥ 3, showing that c(p) is strictly smaller, with the improvement becoming more pronounced for larger p. This will substantiate the claimed improvement over the Carlen-Frank-Lieb exponent. revision: yes
Circularity Check
No significant circularity; derivations are self-contained from explicit constructions and standard Lp estimates
full rationale
The paper's central claims rest on an explicit counterexample construction disproving the c=2 case for p>2, a necessity proof that any valid exponent must satisfy c≤p', a sufficiency proof establishing the inequality at the critical c=p' for integer p≥2, and a three-function bound whose optimal exponent c(p) is obtained via logarithmic optimization over the orthogonality parameter Γ. These steps rely on direct Lp-norm manipulations, product-norm definitions of α_jk, and explicit function choices rather than any reduction to fitted parameters, self-definitional loops, or load-bearing self-citations. The abstract and described results introduce no ansatz smuggled via prior work, no renaming of known patterns as new theorems, and no uniqueness theorems imported from the authors' own earlier papers. The derivation chain is therefore independent of its inputs and self-contained against external Lp-space benchmarks.
Axiom & Free-Parameter Ledger
axioms (2)
- standard math Lp spaces are Banach spaces with the usual norm properties for p >= 2
- domain assumption The product f_j f_k belongs to L^{p/2} so that alpha_jk is well-defined
invented entities (1)
-
Gamma(f1,f2,f3) in [0,1]
no independent evidence
Reference graph
Works this paper leans on
-
[1]
Almost-orthogonality in the Schatten–von Neumann classes.J
Anthony Carbery. Almost-orthogonality in the Schatten–von Neumann classes.J. Operator Theory, 62(1):151–158, 2009
2009
-
[2]
Carlen, Rupert L
Eric A. Carlen, Rupert L. Frank, Paata Ivanisvili, and Elliott H. Lieb. Inequalities for Lp-norms that sharpen the triangle inequality and complement Hanner’s inequality.The Journal of Geometric Analysis, 31:4051–4073, 2021
2021
-
[3]
Carlen, Rupert L
Eric A. Carlen, Rupert L. Frank, and Elliott H. Lieb. Inequalities that sharpen the triangle inequality for sums ofNfunctions inL p.Arkiv f¨ or Matematik, 58(1):57–69, 2020
2020
-
[4]
Sharpening the triangle inequality: envelopes between L2 andL p spaces.Analysis & PDE, 13(5):1591–1603, 2020
Paata Ivanisvili and Connor Mooney. Sharpening the triangle inequality: envelopes between L2 andL p spaces.Analysis & PDE, 13(5):1591–1603, 2020. 39
2020
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.