Random Constructions for Sharp Estimates of Mizohata-Takeuchi Type

Siddharth Mulherkar

REVIEW 3 major objections 5 minor 2 cited by

Random Constructions for Sharp Estimates of Mizohata-Takeuchi Type

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Randomly selected unit balls in a large ball typically give weights that satisfy the sharp Mizohata–Takeuchi estimate up to an R^epsilon loss.

desk verdict New and promising chaining construction for random MT weights, but the main theorem rests on an unproved and currently false comparability in Proposition 27. read the letter →

arxiv 2506.05624 v1 pith:U4THKRAB submitted 2025-06-05 math.CA math.FAmath.PR

Siddharth Mulherkar This is my paper · ORCID

classification math.CAmath.FAmath.PR MSC 42B1042B3760E1560G50

keywords Mizohata–TakeuchiconjectureweightedFourierrestrictionrandomweightschainingmethodBernoulliselectorprocessextensionoperatortubeoccupancycoveringnumbers

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

This paper constructs random weights on a large ball and proves that, with high probability, they satisfy the Mizohata–Takeuchi weighted Fourier restriction inequality up to an R^epsilon loss. The weights are made by independently keeping each unit cell with probability about 1/R, which yields mass about $R^{{d-1}}$ and tube occupancy at most R^epsilon. Such weights are exactly the 'large mass, low tube occupancy' cases where the conjecture is hardest. If the proof is correct, it shows that near-sharp Mizohata–Takeuchi estimates are typical for a simple random model, not a rare delicately engineered phenomenon. The same conclusion is transferred to the alternate model where $R^{{d-1}}$ unit cells are chosen uniformly without replacement.

What carries the argument

The argument treats the supremum over the unit ball of $L^{2}$(Sigma) of the random sum sum_k delta_k integral_{alpha_k} |Eg(x)|^2 dx as the expected supremum of a Bernoulli stochastic process. Chaining controls this expected supremum through covering numbers of the $L^{2}$ unit ball under the seminorm ||g||_~ = max_k |integral_{alpha_k} Eg dx|; the covering numbers come from the empirical method and a duality-of-entropy result, while the concentration estimates come from a Bennett-type large-deviation inequality. A weaker bound with averages |integral_{alpha_k} Eg dx|^2 is upgraded to the full expression integral_{alpha_k} |Eg(x)|^2 dx by splitting each unit cell into about R^epsilon subcells of radius $R^{{-epsilon}}$, Taylor expanding the phase of Eg about subcell centers, and re-running the chaining argument on the resulting monomial-multiplier extension operators Eg_{i_1...i_l}. The Agmon–Hormander trace inequality supplies the baseline unweighted $L^{2}$(B_R) bound that anchors the covering and local-constancy steps.

What would settle it

Check the displayed comparability in the proof of Proposition 27: for a cell $\alpha$' of radius $R^{{-epsilon}}$, the two quantities |integral_{$\alpha$'} Eg dx|^2 and integral_{$\alpha$'} |Eg|^2 dx differ by the factor |$\alpha$'| = $R^{{-epsilon d}}$. Taking g to be any function for which Eg is roughly constant on $\alpha$' makes the discrepancy explicit. A direct re-derivation of the chaining bound for the monomial-multiplier process on $R^{{-epsilon}}$-cells, or a numerical simulation of E sup_{||g||_2 leq 1} sum_k delta_k integral_{$\alpha$'_k} |Eg_{i_1...i_l}(x'_k)|^2 dx at moderate R, would either confirm the claimed R^epsilon bound or exhibit a counterexample that invalidates the main theorem.

Watch

Extended reading notes

Core claim

The paper claims that there exists a random weight w supported in the ball B_R, built as an i.i.d. Bernoulli selection of unit cells with success probability O(1/R), such that with probability at least 1/2 the following hold simultaneously: the weighted $L^{2}$ estimate integral_{B_R} |Eg(x)|^2 w(x) dx is at most C_{epsilon,d} R^epsilon times the $L^{2}$ norm squared of g on Sigma, uniformly over all g in $L^{2}$(Sigma); every 1-tube T has w(T) at most C_epsilon R^epsilon; and the total mass ||w||_1 is comparable to $R^{{d-1}}$. This is a sharp form of the local Mizohata–Takeuchi conjecture for this class of weights, up to the epsilon loss that a known logarithmic counterexample shows cannot be fully removed. The same estimate is proved for the weight model obtained by choosing $R^{{d-1}}$ unit cells uniformly at random without replacement.

Load-bearing premise

The main $L^{2}$ upgrade depends on the claim, sketched rather than fully proved, that the chaining estimate can be rerun after shrinking the unit cells to radius $R^{{-epsilon}}$ and inserting monomial factors into the extension operator; the justification says the proof is the same as for the earlier proposition, and the text also contains a comparability of averages that is off by the factor $R^{{-epsilon d}}$. If that step fails, the random weights may not in fact satisfy the sharp inequality.

Editorial extensions

If this is right

For every smooth compact hypersurface, there is a large class of random weights satisfying the Mizohata–Takeuchi inequality up to R^epsilon while simultaneously having mass R^{d-1} and tube peak R^epsilon.
The uniform-without-replacement model of weights—summing R^{d-1} randomly chosen unit balls—also satisfies the same near-sharp weighted restriction bound with high probability.
The epsilon loss cannot in general be removed, because of the known logarithmic counterexample, so the result sits exactly at the conjecturally optimal level for local Mizohata–Takeuchi estimates.
The proof reduces a weighted restriction problem to bounding expected suprema of Bernoulli processes, so improved chaining or concentration inequalities would directly improve the R^epsilon factor for these weights.
The weighted inequality holds uniformly for all g in L^2(Sigma), not just for individual functions, making the random weights a robust test class for the conjecture.

Reading between the lines

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

The Bernoulli parameter could be varied from 1/R to R^{-alpha} to trace a family of weights interpolating between mass R^{d-1} and smaller masses; the proof suggests a corresponding family of MT-type bounds whose exponents may phase-transition, and this is a natural testable extension.
If generic random weights already achieve the near-sharp estimate, then any counterexample to the local Mizohata–Takeuchi conjecture would need to be substantially more structured than a random large-mass weight, a useful qualitative constraint on where to look.
The subcell decomposition with monomial multipliers is a transferable device: the same local-constancy upgrade could be tried for other random restriction settings, such as discrete extension operators or extension operators for the cone, wherever a similar pointwise Taylor bound on small balls holds.
One could numerically test the claimed bound at moderate R and small epsilon for the monomial-multiplier processes, directly probing whether the sketched Proposition 27 is valid in practice before attempting a full proof.

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

You should know two things about this paper. First, the central idea---constructing weights by i.i.d. Bernoulli selectors and using Talagrand's chaining with Maurey's empirical method to prove they satisfy a local Mizohata-Takeuchi estimate---is genuinely new and worth taking seriously. Second, the main theorem is not proved as written, because the step that upgrades an estimate for averages over unit balls to the full L2 integral on small balls (Proposition 27) is only sketched and contains a false comparability.

What is good: Proposition 12, the weaker statement for |∫_{α_k} Eg|^2, is a real contribution. The covering-number machinery via Maurey and Artstein-Milman-Szarek is well chosen, and the dyadic/chaining argument is recognizable and mostly coherent. The probabilistic estimates in Section 4 (tube occupancy and mass) are fine, and the comparison with Carbery's random weight model shows the method has some flexibility. If Proposition 12 stands, it is already a useful result.

The problem is the hinge. Proposition 27 is exactly what turns the average estimate into the L2 estimate on R^{-ε} balls. The proof says "one can control this expression in the same way one proves Proposition 12" and gives three bullets. That is not a proof: the covering estimate and the concentration need to be redone for the monomial point-evaluation functionals α'_k, and the deterministic bound in Lemma 35 carries an R^{εd} loss that has to be absorbed. More seriously, the comparability claimed at the top of the proof is wrong: |∫_{α'} Eg dx|^2 is R^{-2εd}|Eg|^2, while ∫_{α'}|Eg|^2 dx is R^{-εd}|Eg|^2; the ratio is the volume of α', not a constant. The author's R^{-ε} rescaling of B might absorb an R^ε loss, but as written the factor is R^{εd}, and the assertion is simply not true. There are also smaller defects: broken cross-references to "Proposition 40" and a few notational slips in Section 2. These alone would be minor; the Proposition 27 problem is not.

Who this is for: people working on weighted restriction estimates and the local Mizohata-Takeuchi conjecture. The chaining perspective is likely to be useful even if this particular theorem needs repair. The result is plausible, and the author has identified the right tools, but a referee would need to see a complete proof of Proposition 27 (or a different route) before the main claim can be trusted. I would send it to peer review---a serious referee can engage with the chaining framework, and the gap is concrete---but I would not cite the main theorem in its current form.

My recommendation: invite revision, with Proposition 27 as the explicit focus.

Referee Report

3 major / 5 minor

Summary. The paper constructs random weights supported on the ball B_R by independently selecting unit cubes with probability O(1/R) and claims that, with high probability, such weights have mass comparable to R^{d-1}, tube occupancy at most R^ε, and satisfy the localized Mizohata–Takeuchi inequality ∫ |Eg|^2 w ≲_ε R^ε ∫ |g|^2 dσ. The main technical step is an expectation bound (Theorem 9 / Proposition 10) for the supremum over the L^2(Σ) unit ball of a random weighted sum of local L^2 norms of the extension operator. Section 2 proves a weaker averaged version using Maurey's empirical covering method, the Artstein–Milman–Szarek entropy duality, and a Talagrand-type chaining argument with Bennett concentration. Section 3 attempts to upgrade to the full L^2 bound via local constancy and Taylor expansion, relying on Proposition 27 for point evaluations of monomial extension operators. Section 4 proves the mass and tube-occupancy properties and extends the result to Carbery's random weight model.

Significance. If the proof can be completed, the result would be a genuine contribution: it exhibits a large class of weights satisfying the local Mizohata–Takeuchi estimate up to an R^ε loss, complementing Cairo's recent counterexample that shows a log R loss is necessary, and it introduces techniques from high-dimensional probability into restriction theory in a novel way. The paper is not circular: it uses the Agmon–Hörmander trace inequality (the MT estimate for the weight 1_{B_R}), Maurey's empirical method, the Artstein–Milman–Szarek duality, and Talagrand–Bennett concentration, none of which reduce to the desired conclusion by definition. The informal exposition and the appendix are carefully written and useful. The main weakness is that the proof of Proposition 27, which is the hinge of the entire upgrade, is only sketched and contains a false comparability statement; therefore the central claim is not yet established as written, although the strategy appears plausible and repairable.

major comments (3)

[Section 3, Proposition 27] The displayed comparability |∫_{α'_k} Eg_{i_1...i_l}(x'_k) dx|^2 ∼ R^{-εd}|Eg(x'_k)|^2 ∼ ∫_{α'_k} |Eg(x'_k)|^2 dx is not correct: the first quantity equals R^{-2εd}|Eg|^2 and the second equals R^{-εd}|Eg|^2, so the two sides differ by the volume factor |α'_k| = R^{-εd}. This matters because the proof of (40) uses this equivalence to reduce the L^2 integral over the small ball to the square of an average; a corrected argument must handle the missing factor, for instance by absorbing it into the final R^ε loss after redefining ε, but that step is not written.
[Section 3, Proposition 27] The proof of Proposition 27 is a sketch rather than a complete rerun of Proposition 12. The sentence 'one can control this expression in the same way one proves Proposition 12' leaves unproved four load-bearing steps: (i) the covering-number bound for the new dual functionals \tilde{α}'_k(ω) = ∫_{α'_k} e^{2πiω·x'_k} ω_{i_1}...ω_{i_l} dx, which requires the correct norm bound ∥\tilde{α}'_k∥_{L^2(Σ)} = R^{-εd} (not O(1) as stated in bullet (1)); (ii) the deterministic estimate Σ_k |E g_{i_1...i_l}(x'_k)|^2 ≲ R^{1+εd} from Lemma 35 and its analogue of (20); (iii) the propagation of the R^{εd} factor through the Bennett-type concentration and the dyadic decomposition, with O(ε log R) active scales; and (iv) the conversion of the resulting bound into the claimed R^ε estimate. Because Proposition 27 is precisely the step that upgrades the unit-ball averages of Section 2 to the L^2 integrals over R^{-ε}-balls that appear in Proposition 10, Theorems 9 and 6(1) are not established as written. The author should supply the complete argument or clearly state it as a conjecture.
[Section 2, Lemma 22] The inequality (26) asserts 2δ card(I(g)) ≤ 2^{2k+3}/log R, but combining (20) with δ = O(1/R) only gives 2δ card(I(g)) ≲ 2^{2k+3}; the extra factor 1/log R would require the sum in (20) to be O(R/log R), which is not a consequence of the Agmon–Hörmander bound. A corrected version of Lemma 22 is needed; the proof likely survives with a dimensional constant in place of 1/log R, but as written the displayed bound is false.

minor comments (5)

[Section 1.3] After Theorem 9, the sentence 'which proves property 3' should read 'property 1'; the mass property (property 3) is proved later by Lemma 28.
[Section 2, equation (24) and surrounding text] The indices k and j are interchanged in several sums in the dyadic decomposition (for example, '∑_{j=1}^n δ_j l_{j,k}(g)' and later '∑_{j=1}^n ∫_{α_k} |Eg|^2 dx'); please standardize the notation.
[Section 2, Remark 24] The sentence 'By repeating the same argument with sin ... will show that (15) implies (16)' appears to have the implication direction reversed; repeating the argument with sin gives the complex bound (15), not the reverse.
[Section 2, proof of Proposition 18] The phrase 'along with (18)' should refer to the Agmon–Hörmander bound or to Proposition 18 itself; the cross-reference is unclear.
[Throughout] There are several typographical and cross-referencing errors: Theorem 2 displays R^{(d+1)/(d-1)+ε} where the text intends R^{(d-1)/(d+1)+ε}; Proposition 27 is referred to as 'Proposition 40' in the proof of Proposition 10; and Lemma 22 cites 'Corollary 22' and 'Corollary 8' instead of the intended corollary.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the random-weight Mizohata-Takeuchi estimate is derived from independent external tools (Agmon-Hormander trace inequality, Maurey's empirical method, Artstein-Milman-Szarek entropy duality, Talagrand-Bennett concentration). The only notable weakness is an incompletely proved Proposition 27, which is a correctness gap, not a circular reduction.

full rationale

The derivation chain is not circular. The weak bound in Proposition 12 is proved by combining Agmon-Hormander trace inequality for the fixed weight 1_{B_R} (Proposition 18), Maurey's empirical covering argument (Theorem 13), the Artstein-Milman-Szarek duality theorem (Theorem 14), and Bennett/Talagrand concentration (Proposition 19). None of these inputs equals the target random-weight MT estimate: Agmon-Hormander gives only the classical bound R ||g||_2^2 for the full unit-ball weight, which is not the sparse random weight conclusion. Proposition 10 upgrades to full L^2 averages via local constancy (Lemma 25), a rerun of the chaining argument with monomial multipliers (Proposition 27), and the separated-point trace inequality (Lemma 35). Lemma 35 itself is derived from the independent Agmon-Hormander and Tomas-Stein inequalities, not from the theorem being proved. The tube-occupancy bound (Lemma 28) is a separate Chernoff large-deviation estimate, and the Carbery-model comparison (Proposition 29) invokes the already-established Proposition 10. There are no fitted parameters, no self-citations, and no imported uniqueness theorem that forces the conclusion. The manuscript's real weakness is that Proposition 27 is only sketched ('one can control this expression in the same way one proves Proposition 12') and its written comparability between squared averages and L^2 integrals is dimensionally off by a factor |alpha'| = R^{-epsilon d}. That is a completeness/correctness risk in the proof as written, but it does not make the argument circular: the missing work is a technical rerun of the same independent chaining scheme, not an appeal to the desired conclusion by definition.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

The proof is built on standard results in restriction theory and high-dimensional probability. The only tuned object is the Bernoulli success probability delta, which sets the mass scale. No new physical or mathematical entities are postulated.

free parameters (1)

delta (selector probability) = c/R with an absolute constant c
Chosen by hand so that E||w||_1 ~ R^{d-1} and sup_T w(T) stays small. Not fitted to data, but a design parameter of the random model.

assumptions (6)

standard math Agmon-Hormander trace inequality: integral over B_R of |Eg|^2 dx is bounded by C R integral over Sigma of |g|^2 dsigma
Used as the baseline MT-type estimate for w = 1_{B_R}; proved in the appendix.
standard math Maurey's empirical method for covering numbers of convex hulls
Theorem 13; used to control covering numbers of the hull of the test functions.
standard math Artstein-Milman-Szarek duality of metric entropy
Theorem 14; converts covering of the convex hull into covering of the dual unit ball.
standard math Bennett-type large deviation inequality
Proposition 19/31; gives concentration for sums of bounded zero-mean variables.
standard math Tomas-Stein restriction theorem
Used in Lemma 35 to control sums over separated points of |Eg|^2.
domain assumption Sigma is a smooth compact hypersurface with surface measure; weights supported in B_R
Standard setting for the local Mizohata-Takeuchi problem; no curvature condition is imposed in the main theorem.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Random Constructions for Sharp Estimates of Mizohata-Takeuchi Type." pith.science (2026). https://pith.science/paper/U4THKRAB

@misc{pith2026250605624,
  author       = {Pith},
  title        = {Pith review of: Random Constructions for Sharp Estimates of Mizohata-Takeuchi Type},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U4THKRAB}},
  note         = {Machine review of arXiv:2506.05624}
}

read the original abstract

A Mizohata-Takeuchi type estimate is a type of weighted Fourier restriction estimate. Using tools from high dimensional probability, we construct a large class of weights that satisfy sharp estimates of Mizohata-Takeuchi type. One can interpret our result as saying that with high probability, a generic weight satisfies a sharp inequality of Mizohata-Takeuchi type (up to an epsilon-loss).

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

Lebesgue measure of distance sets with regular pins and multi-scale Mizohata-Takeuchi-type estimates
math.CA 2026-03 unverdicted novelty 8.0 of 10

Under dim_H E >1, dim_H E + dim_H F >2 and F regular (equal Hausdorff and packing dimensions), there exists y in F such that the pinned distance set Δ_y(E) has positive Lebesgue measure.
Mizohata-Takeuchi inequalities for orthonormal systems
math.CA 2025-06 conditional novelty 8.0 of 10

Orthonormal systems of inputs satisfy a global Mizohata-Takeuchi-type weighted inequality for the sphere and paraboloid, with an X-ray transform norm over the midpoint set K^diamond on the right-hand side.

Reference graph

Works this paper leans on

22 extracted references · 20 canonical work pages · cited by 2 Pith papers

[1]

Milman, and Stanislaw J

Shiri Artstein, Vitali D. Milman, and Stanislaw J. Szarek. Duality of metric entropy. Annals of Mathematics, 159(3):1313–1328, 2004

work page 2004
[2]

A stein conjecture for the circle

Jonathan Bennett, Anthony Carbery, Fernando Soria, and Ana Vargas. A stein conjecture for the circle. Mathematische Annalen, 336(3):671–695, 2006

work page 2006
[3]

Bourgain, A

J. Bourgain, A. Pajor, S. J. Szarek, and N. Tomczak-Jaegermann. On the duality problem for entropy numbers of operators, pages 50–63. Springer Berlin Heidelberg, Berlin, Heidelberg, 1989

work page 1989
[4]

Bounded orthogonal systems and the λ(p)-set problem

Jean Bourgain. Bounded orthogonal systems and the λ(p)-set problem. Acta Mathematica, 162(3– 4):227–245, 1989

work page 1989
[5]

A counterexample to the mizohata-takeuchi conjecture

Hannah Mira Cairo. A counterexample to the mizohata-takeuchi conjecture. arXiv preprint arXiv:2502.06137, 2025. Submitted on February 10, 2025

arXiv 2025
[6]

Large sets with limited tube occupancy

Anthony Carbery. Large sets with limited tube occupancy. Journal of the London Mathematical Society, 79(2):529–543, 2009

work page 2009
[7]

The mizohata–takeuchi conjecture: An overview, 2022

Anthony Carbery. The mizohata–takeuchi conjecture: An overview, 2022. Accessed: 2025-05-23

work page 2022
[8]

H¨ anninen, and Stef´ an Ingi Valdimarsson

Anthony Carbery, Timo S. H¨ anninen, and Stef´ an Ingi Valdimarsson. Disentanglement, multilinear duality and factorisation for nonpositive operators. Analysis & PDE , 16(2):511–543, 2023

work page 2023

Show all 22 references

[9]

Some sharp inequalities of mizohata–takeuchi-type

Anthony Carbery, Marina Iliopoulou, and Hong Wang. Some sharp inequalities of mizohata–takeuchi-type. Revista Matem´ atica Iberoamericana, 40(4):1387–1418, 2024

2024
[10]

On the n-set occupancy problem

Ciprian Demeter and Ruixiang Zhang. On the n-set occupancy problem. arXiv preprint arXiv:2403.10678, 2024

2024 arXiv
[11]

New improvement to falconer’s distance set problem in higher dimensions

Xiumin Du, Yumeng Ou, Kevin Ren, and Ruixiang Zhang. New improvement to falconer’s distance set problem in higher dimensions. arXiv preprint arXiv:2309.04103 , 2023

2023 arXiv
[12]

Sharp l2 estimate of schr¨ odinger maximal function in higher dimensions

Xiumin Du and Ruixiang Zhang. Sharp l2 estimate of schr¨ odinger maximal function in higher dimensions. Annals of Mathematics , 189(3):837–861, 2019

2019
[13]

An enemy scenario in restriction theory

Larry Guth. An enemy scenario in restriction theory. https://www.youtube.com/watch?v= x-DET83UjFg, 2022. Accessed: 2025-05-21

2022
[14]

On falconer’s distance set problem in the plane

Larry Guth, Alex Iosevich, Yumeng Ou, and Hong Wang. On falconer’s distance set problem in the plane. Inventiones Mathematicae, 219(3):779–830, 2020

2020
[15]

On the Cauchy Problem, volume 3 of Notes and Reports in Mathematics in Science and Engineering

Sigeru Mizohata. On the Cauchy Problem, volume 3 of Notes and Reports in Mathematics in Science and Engineering. Academic Press, 1985. Edited by William F. Ames

1985
[16]

A sharp mizohata–takeuchi type estimate for the cone in R3

Alex Ortiz. A sharp mizohata–takeuchi type estimate for the cone in R3. arXiv preprint arXiv:2307.11731, 2023

2023
[17]

A necessary condition for the well-posedness of the cauchy problem for a certain class of evolution equations

Jiro Takeuchi. A necessary condition for the well-posedness of the cauchy problem for a certain class of evolution equations. Proceedings of the Japan Academy, 50(2):133–137, 1974

1974
[18]

On the cauchy problem for some non-kowalewskian equations with distinct charac- teristic roots

Jiro Takeuchi. On the cauchy problem for some non-kowalewskian equations with distinct charac- teristic roots. Journal of Mathematics of Kyoto University , 20(1):105–124, 1980

1980
[19]

Constructions of majorizing measures, bernoulli processes and cotype

Michel Talagrand. Constructions of majorizing measures, bernoulli processes and cotype. Geometric and Functional Analysis , 4(6):660–717, 1994

1994
[20]

Sections of smooth convex bodies via majorizing measures

Michel Talagrand. Sections of smooth convex bodies via majorizing measures. Acta Mathematica, 175(2):273–300, 1995

1995
[21]

Upper and Lower Bounds for Stochastic Processes: Modern Methods and Classical Problems, volume 60 of Ergebnisse der Mathematik und ihrer Grenzgebiete

Michel Talagrand. Upper and Lower Bounds for Stochastic Processes: Modern Methods and Classical Problems, volume 60 of Ergebnisse der Mathematik und ihrer Grenzgebiete. 3. Folge / A Series of Modern Surveys in Mathematics . Springer, 2014

2014
[22]

High-Dimensional Probability: An Introduction with Applications in Data Sci- ence, volume 47 of Cambridge Series in Statistical and Probabilistic Mathematics

Roman Vershynin. High-Dimensional Probability: An Introduction with Applications in Data Sci- ence, volume 47 of Cambridge Series in Statistical and Probabilistic Mathematics . Cambridge Uni- versity Press, Cambridge, 2018. Department of Mathematics, University of California, ...

2018

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Milman, and Stanislaw J

Shiri Artstein, Vitali D. Milman, and Stanislaw J. Szarek. Duality of metric entropy. Annals of Mathematics, 159(3):1313–1328, 2004

work page 2004

[2] [2]

A stein conjecture for the circle

Jonathan Bennett, Anthony Carbery, Fernando Soria, and Ana Vargas. A stein conjecture for the circle. Mathematische Annalen, 336(3):671–695, 2006

work page 2006

[3] [3]

Bourgain, A

J. Bourgain, A. Pajor, S. J. Szarek, and N. Tomczak-Jaegermann. On the duality problem for entropy numbers of operators, pages 50–63. Springer Berlin Heidelberg, Berlin, Heidelberg, 1989

work page 1989

[4] [4]

Bounded orthogonal systems and the λ(p)-set problem

Jean Bourgain. Bounded orthogonal systems and the λ(p)-set problem. Acta Mathematica, 162(3– 4):227–245, 1989

work page 1989

[5] [5]

A counterexample to the mizohata-takeuchi conjecture

Hannah Mira Cairo. A counterexample to the mizohata-takeuchi conjecture. arXiv preprint arXiv:2502.06137, 2025. Submitted on February 10, 2025

arXiv 2025

[6] [6]

Large sets with limited tube occupancy

Anthony Carbery. Large sets with limited tube occupancy. Journal of the London Mathematical Society, 79(2):529–543, 2009

work page 2009

[7] [7]

The mizohata–takeuchi conjecture: An overview, 2022

Anthony Carbery. The mizohata–takeuchi conjecture: An overview, 2022. Accessed: 2025-05-23

work page 2022

[8] [8]

H¨ anninen, and Stef´ an Ingi Valdimarsson

Anthony Carbery, Timo S. H¨ anninen, and Stef´ an Ingi Valdimarsson. Disentanglement, multilinear duality and factorisation for nonpositive operators. Analysis & PDE , 16(2):511–543, 2023

work page 2023