arxiv: 2604.14594 · v2 · submitted 2026-04-16 · ✦ hep-lat · hep-ph· hep-th

Recognition: unknown

An efficient Wavelet-Based Hamiltonian Formulation of Quantum Field Theories using Flow-Equations

Mrinmoy Basak , Debsubhra Chakraborty , Nilmani Mathur

Authors on Pith no claims yet

Pith reviewed 2026-05-10 10:32 UTC · model grok-4.3

classification ✦ hep-lat hep-phhep-th

keywords wavelet basissimilarity renormalization groupHamiltonian formulationquantum field theorylow-energy spectrumDaubechies waveletsfree scalar fieldcomputational efficiency

0 comments

The pith

A wavelet basis plus similarity renormalization group flow equations lets low-energy spectra be read off from the smallest resolution block of the Hamiltonian.

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

The paper shows how to rewrite a quantum field theory Hamiltonian in a Daubechies wavelet basis so that the degrees of freedom naturally separate into blocks labeled by spatial resolution. Similarity renormalization group flow equations are then used to remove the coupling between blocks at different resolutions. Once the blocks are decoupled, the low-lying energies of the full theory are recovered simply by diagonalizing the lowest-resolution block, and this approximation improves as the overall resolution cutoff is raised. The result is a concrete reduction in matrix size and therefore in the computational cost of spectrum calculations, demonstrated explicitly for the free scalar field.

Core claim

When the free scalar field is expanded in a Daubechies wavelet basis the Hamiltonian becomes a collection of fixed-resolution blocks plus inter-resolution couplings; the similarity renormalization group flow equations drive those inter-resolution couplings to zero while leaving the intra-block matrix elements of the lowest-resolution block essentially unchanged, so that the eigenvalues of this smallest block converge to the true low-energy spectrum of the original theory.

What carries the argument

Daubechies wavelet basis that converts the field theory into coupled localized oscillators organized by resolution index, combined with SRG flow equations that systematically decouple the resolution blocks.

If this is right

Low-energy spectrum calculations require only the smallest block rather than the full high-resolution matrix.
Computational cost decreases as the maximum resolution is increased because only the lowest-resolution block needs to be diagonalized.
The method supplies a systematic, non-perturbative way to truncate the Hilbert space while keeping the infrared physics intact.
The same wavelet-plus-flow construction can be applied to any Hamiltonian that admits a position-space wavelet expansion.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The approach may allow Hamiltonian-based real-time evolution studies on systems too large for conventional basis expansions.
Because the wavelets are localized, the same framework could be adapted to lattice gauge theories or models with local interactions.
The block-decoupling property suggests a direct link to traditional renormalization-group ideas, offering a Hamiltonian route to scale separation.

Load-bearing premise

The similarity renormalization group flow successfully removes coupling between different resolution blocks while preserving the low-energy eigenvalues inside the lowest-resolution block without introducing large uncontrolled errors.

What would settle it

Direct numerical comparison, at successively higher wavelet resolutions, between the low-lying eigenvalues obtained from the full unflowed Hamiltonian and those obtained from the flowed lowest-resolution block alone; persistent disagreement would falsify the claim.

Figures

Figures reproduced from arXiv: 2604.14594 by Debsubhra Chakraborty, Mrinmoy Basak, Nilmani Mathur.

read the original abstract

We propose an effective Hamiltonian formulation of quantum field theories using a Daubechies wavelet basis in position space. Combined with flow-equation methods of the similarity renormalization group (SRG), this approach provides an efficient framework for analyzing quantum field theories by reducing the dimensionality of the Hamiltonian and systematically decoupling degrees of freedom across scales. As an application, the free scalar field theory has been reformulated within this framework to calculate the low-lying energy spectrum of the theory. These basis elements are known to transform the free scalar field theory into a theory of coupled localized oscillators, each of which is labeled by a location and a resolution index. In this representation, the Hamiltonian is naturally organized into fixed-resolution blocks, alongside blocks associated with the interactions between different resolutions. To decouple the different resolution modes and obtain a block diagonalized Hamiltonian with each block associated with a fixed resolution, the flow equation approach of SRG is applied. Finally, we demonstrate that with increasing resolution, the low-energy spectrum can be extracted from the effective lowest-resolution block of the Hamiltonian, leading to a significant reduction in computational cost.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Wavelet basis plus SRG flow gives a plausible way to block-diagonalize Hamiltonian QFTs, but the cost savings still need to be shown against the full-matrix requirement of the flow.

read the letter

The paper's core move is to put the free scalar field on a Daubechies wavelet basis so the Hamiltonian breaks into fixed-resolution blocks plus cross terms, then run SRG flow to decouple the resolutions and read the low-energy spectrum off the coarsest block alone. That combination is new enough to be worth noticing. The wavelet part gives a clean position-space localization with resolution labels, and SRG is a standard tool for decoupling, so the pairing makes sense for trying to shrink the effective matrix size. They show the block structure explicitly for the free theory and apply the flow. If the numerics back it up, this could be a practical route for Hamiltonian calculations where you want to focus on infrared modes without keeping all ultraviolet degrees of freedom. The main question mark is whether the claimed cost reduction survives the flow step itself. SRG evolution involves repeated operations on the whole matrix, so unless they have a way to truncate or sparsify during the integration, you might still need the full high-resolution object to start with. The abstract does not spell out any such shortcut or report run times, so the savings remain unproven on the page. There are also no sample spectra, convergence checks, or comparisons to the exact free-field energies, which leaves the practical performance open. This work is aimed at lattice field theorists who already use Hamiltonian methods or are exploring multi-scale bases. A reader looking for new tools to tame large matrices in QFT Hamiltonians would get something concrete to try. I would send it for peer review. The technical setup is clear and the test case is simple enough that referees can check the claims directly.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a Daubechies wavelet basis in position space for reformulating QFT Hamiltonians, combined with SRG flow equations to decouple fixed-resolution blocks. For the free scalar field, it claims that after flow the low-energy spectrum is recoverable from the lowest-resolution block alone, yielding substantial computational savings as resolution increases.

Significance. If the SRG flow can be performed without retaining or operating on the full high-resolution matrix while still preserving the low-energy physics in the coarse block, the wavelet-plus-SRG combination would offer a systematic multi-resolution approach to Hamiltonian QFTs. The localization of oscillators by wavelets is a conceptually attractive feature, but the lack of any numerical spectra, convergence data, or error quantification for the free scalar case leaves the practical significance unestablished.

major comments (2)

Abstract and the demonstration paragraph: the central claim that 'with increasing resolution, the low-energy spectrum can be extracted from the effective lowest-resolution block' is asserted without any reported energy eigenvalues, comparison to the exact free-scalar dispersion, error measures, or plots of convergence versus resolution; this absence directly undermines the stated reduction in computational cost.
Section describing the SRG implementation (around the flow-equation application): the standard generator η = [H_d, H] and the differential equation dH/ds = [η, H] require repeated full-matrix multiplications; the manuscript provides no reduced-subspace algorithm, sparsity exploitation, or block-wise integration scheme that would avoid storing and updating the high-resolution blocks during the flow, which is load-bearing for the efficiency claim.

minor comments (2)

The mapping from the continuum free scalar to the coupled-oscillator Hamiltonian in the wavelet basis would benefit from an explicit low-dimensional example (e.g., two or three resolution levels) showing the block structure before and after flow.
Notation for the resolution index and the block partitioning of the Hamiltonian matrix should be introduced with a small diagram or table to make the decoupling statement easier to follow.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful and constructive review of our manuscript. The comments correctly identify areas where additional evidence and clarification are needed to support the central claims. We provide point-by-point responses below and commit to revisions that strengthen the work without altering its core framework.

read point-by-point responses

Referee: Abstract and the demonstration paragraph: the central claim that 'with increasing resolution, the low-energy spectrum can be extracted from the effective lowest-resolution block' is asserted without any reported energy eigenvalues, comparison to the exact free-scalar dispersion, error measures, or plots of convergence versus resolution; this absence directly undermines the stated reduction in computational cost.

Authors: We agree that the manuscript would be substantially strengthened by explicit numerical support for the claim. The theoretical structure—wavelet transformation to localized oscillators followed by SRG decoupling into resolution blocks—is derived in detail, yet we did not include concrete eigenvalue extractions, comparisons to the analytic dispersion relation, or convergence diagnostics in the submitted version. In the revised manuscript we will add numerical results for the free scalar field at multiple resolutions, direct comparisons of the extracted low-lying eigenvalues to the exact continuum dispersion, quantitative error measures, and plots of spectral convergence versus resolution. These additions will directly substantiate the asserted computational savings. revision: yes
Referee: Section describing the SRG implementation (around the flow-equation application): the standard generator η = [H_d, H] and the differential equation dH/s = [η, H] require repeated full-matrix multiplications; the manuscript provides no reduced-subspace algorithm, sparsity exploitation, or block-wise integration scheme that would avoid storing and updating the high-resolution blocks during the flow, which is load-bearing for the efficiency claim.

Authors: The referee accurately notes that the conventional SRG flow equations involve full-matrix operations. Our current implementation applies the flow to the complete wavelet-basis Hamiltonian, which is block-structured by resolution but is not yet integrated with a reduced-subspace or block-decoupled integrator. The primary efficiency demonstrated in the work is the post-flow extraction of low-energy physics from the coarsest block alone, avoiding full diagonalization of the high-resolution matrix. For the modest system sizes used in the proof-of-principle calculations this is feasible, but we recognize that scalability to higher resolutions requires further optimization. In the revision we will (i) report the observed computational scaling of the present implementation, (ii) discuss how the resolution-block structure can be exploited to develop a block-wise flow scheme, and (iii) either implement a prototype of such a scheme or moderate the efficiency claims to reflect the current scope. This constitutes a partial revision that addresses the concern while preserving the conceptual contribution. revision: partial

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper applies established Daubechies wavelet bases and SRG flow-equation methods to reformulate the free scalar field Hamiltonian into resolution-blocked form, then demonstrates that SRG decoupling allows low-energy eigenvalues to be read from the coarsest block alone. This is presented as a numerical consequence of the flow rather than a definitional identity or a parameter fit. No equation in the abstract or described chain equates the final spectrum extraction to the input Hamiltonian by construction, nor does any load-bearing step rely on a self-citation whose content is itself unverified. The claimed cost reduction follows from the block-decoupling property of the SRG generator, which is an external mathematical tool, not an internal redefinition. The derivation therefore remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The abstract invokes standard properties of Daubechies wavelets and the SRG flow method but introduces no new free parameters, fitted constants, or postulated entities; all elements are drawn from existing mathematical and renormalization-group literature.

axioms (2)

domain assumption Daubechies wavelets transform the free scalar field into a set of coupled localized oscillators labeled by location and resolution index
Stated directly in the abstract as the starting representation of the theory.
domain assumption SRG flow equations can be applied to remove inter-resolution couplings while preserving low-energy information in the lowest-resolution block
Core premise required for the block-diagonalization and spectrum-extraction claim.

pith-pipeline@v0.9.0 · 5499 in / 1440 out tokens · 60210 ms · 2026-05-10T10:32:43.756952+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

52 extracted references · 7 canonical work pages

[1]

The mother scaling functions(x)and its integer translatess(x−n)are orthonormal, Z s(x)s(x−n)dx=δ 0n,(3) which yields the relation 2K−1X m=0 hm hm−2n =δ 0n.(4)
[2]

The integral of the mother scaling function is normalized to unity, Z s(x)dx= 1,(5) which gives 2K−1X n=0 hn = √ 2.(6)
[3]

(4), (6), and (8), yields two sets of solutions for the coefficients,hl andh ′ l

Any polynomial of degreem(0≤m < K) can be expressed in terms of the mother scaling function and its translates, xm = ∞X n=−∞ cn s(x−n),(7) 3 which leads to 2K−1X n=0 nmgn = 2K−1X n=0 nm(−1)nh2K−1−n = 0, m < K.(8) Solving Eqs. (4), (6), and (8), yields two sets of solutions for the coefficients,hl andh ′ l. These solutions are related byh l =h ′ 2K−l−l, th...
[4]

Consequently, each Fock basis elementcontains10modes, resultinginatotalFockspace dimension of 184,756

Periodic boundary conditions are implemented by extending theΩmatrix through periodic repetition of its elements, such that the 5th element is identified with the 7th, the 4th with the 8th, the 3rd with the 9th, and the 2nd with the 10th. Consequently, each Fock basis elementcontains10modes, resultinginatotalFockspace dimension of 184,756. The Hamiltonian...

1900
[5]

Generate all the raw Fock-space basis states within the resolution-truncated subspace of the Hilbert space
[6]

The representative state is the one that will be used to generate the translation-invariant state

Identify the representative states among them. The representative state is the one that will be used to generate the translation-invariant state. A state is considered a representative state if it is the smallest (according to a chosen ordering) among all its translational copies. For example, consider the raw state|2,3,1,4⟩. This is not the representativ...
[7]

For example, translationally invariant state|ϕ⟩ corresponding to the representative state|1,4,2,3⟩ is, |ϕ⟩ → 1√ 4 (|1,4,2,3⟩+|3,1,4,2⟩+|2,3,1,4⟩+|4,2,3,1⟩)

After identifying the representative state, we form a linear combination of all the distinct translational copies of the state to construct a translation-invariant state that resides within the zero-momentum sector. For example, translationally invariant state|ϕ⟩ corresponding to the representative state|1,4,2,3⟩ is, |ϕ⟩ → 1√ 4 (|1,4,2,3⟩+|3,1,4,2⟩+|2,3,1...
[8]

, nN | ˆT −iH ˆT j |n1,

Calculate the Hamiltonian matrix within this translation-invariant Fock-space basis, which yields the zero-momentum sector of the Hamiltonian in this framework: Hmn =⟨ϕ m|H|ϕ n⟩ = 1√SmSn Sm−1X i=0 Sn−1X j=0 ⟨n1, . . . , nN | ˆT −iH ˆT j |n1, . . . , nN ⟩ = r Sm Sn Sn−1X j=0 ⟨n1, . . . , nN |HT j |n1, . . . , nN ⟩.(D3) Appendix E: Construction of the even ...
[9]

For example, let’s consider|1,0,3,0,1⟩

Symmetric eigenstates (Eigenvalues=+1) If the state is perfectly symmetric around the center of thelattice, theoperationoftheparityoperatorgivesback the same state. For example, let’s consider|1,0,3,0,1⟩. The application of parity operator, ˆP, onto this state returns the same state with eigenvalue+1
[10]

For example, let’s consider a state|2,0,0,0,0⟩

Asymmetric state (Not an eigenstate) InthewaveletbasedFockspacebasismostofthestates fall under this category. For example, let’s consider a state|2,0,0,0,0⟩. The parity operation inverses the state to|0,0,0,0,2⟩. So,|2,0,0,0,0⟩is not an eigenstate of ˆP. Although using these states, we can construct the eigen basses of even parity and the odd parity secto...
[11]

Haar, Zur theorie der orthogonalen funktionensysteme, Mathematische Annalen69, 331 (1910)

A. Haar, Zur theorie der orthogonalen funktionensysteme, Mathematische Annalen69, 331 (1910)

1910
[12]

Getoor, J

R. Getoor, J. l. doob: Foundations of stochastic processes and probabilistic potential theory, The Annals of Probability37, 10.1214/09-AOP465 (2009)

work page doi:10.1214/09-aop465 2009
[13]

Daubechies, Orthonormal bases of compactly supported wavelets, Communications on Pure and Applied Mathematics41, 909 (1988)

I. Daubechies, Orthonormal bases of compactly supported wavelets, Communications on Pure and Applied Mathematics41, 909 (1988)

1988
[14]

Mallat, A theory for multiresolution signal decomposition: the wavelet representation, IEEE Transactions on Pattern Analysis and Machine 17 Intelligence11, 674 (1989)

S. Mallat, A theory for multiresolution signal decomposition: the wavelet representation, IEEE Transactions on Pattern Analysis and Machine 17 Intelligence11, 674 (1989)

1989
[15]

Rev.140, B445 (1965)

K.G.Wilson,Modelhamiltoniansforlocalquantumfield theory, Phys. Rev.140, B445 (1965)

1965
[16]

K. G. Wilson, T. S. Walhout, A. Harindranath, W.-M. Zhang, R. J. Perry, and S. D. Głazek, Nonperturbative qcd: A weak-coupling treatment on the light front, Phys. Rev. D49, 6720 (1994)

1994
[17]

Battle, P

G. Battle, P. Federbush, and P. Uhlig, Wavelets for quantum gravity and divergence-free wavelets, Applied and Computational Harmonic Analysis1, 295 (1994)

1994
[18]

Federbush, A new formulation and regularization of gauge theories using a non-linear wavelet expansion, Progress of Theoretical Physics94, 1135 (1995)

P. Federbush, A new formulation and regularization of gauge theories using a non-linear wavelet expansion, Progress of Theoretical Physics94, 1135 (1995)

1995
[19]

W. N. Polyzou, T. Michlin, and F. Bulut, Multi-scale methods in quantum field theory, Few-Body Systems59, 36 (2018)

2018
[20]

T. L. Michlin,Using wavelet bases to separate scales in quantum field theory(The University of Iowa, 2017)

2017
[21]

C. Best, Wavelet-induced renormalization group for the landau-ginzburg model, Nuclear Physics B - Proceedings Supplements83-84, 848 (2000), proceedings of the XVIIth International Symposium on Lattice Field Theory

2000
[22]

Evenbly and S

G. Evenbly and S. R. White, Entanglement renormalization and wavelets, Phys. Rev. Lett.116, 140403 (2016)

2016
[23]

E. T. Tomboulis, Wavelet field decomposition and uv ‘opaqueness’, Journal of High Energy Physics2021, 77 (2021)

2021
[24]

D. J. George, Y. R. Sanders, M. Bagherimehrab, B. C. Sanders, and G. K. Brennen, Entanglement in quantum field theory via wavelet representations, Phys. Rev. D 106, 036025 (2022)

2022
[25]

Thiemann, Renormalization, wavelets, and the dirichlet-shannon kernels, Phys

T. Thiemann, Renormalization, wavelets, and the dirichlet-shannon kernels, Phys. Rev. D108, 125008 (2023)

2023
[26]

Best and A

C. Best and A. Schaefer, Variational description of statistical field theories using daubechies’ wavelets (1994), arXiv:hep-lat/9402012 [hep-lat]

work page arXiv 1994
[27]

Halliday and P

I. Halliday and P. Suranyi, Simulation of field theories in wavelet representation, Nuclear Physics B436, 414 (1995)

1995
[28]

A. E. Ismail, G. Stephanopoulos, and G. C. Rutledge, Multiresolution analysis in statistical mechanics. ii. the wavelet transform as a basis for monte carlo simulations on lattices, The Journal of Chemical Physics118, 4424 (2003)

2003
[29]

W. N. Polyzou, Path integrals, complex probabilities and the discrete weyl representation, Journal of Physics A: Mathematical and Theoretical57, 275201 (2024)

2024
[30]

Bulut and W

F. Bulut and W. N. Polyzou, Wavelets in field theory, Phys. Rev. D87, 116011 (2013)

2013
[31]

Neuberger, Wavelets and lattice field theory., EPJ Web of Conferences175, 11002 (2018)

H. Neuberger, Wavelets and lattice field theory., EPJ Web of Conferences175, 11002 (2018)

2018
[32]

M. V. Altaisky, Wavelet-based quantum field theory, Symmetry, Integrability and Geometry: Methods and Applications 10.3842/SIGMA.2007.105 (2007)

work page doi:10.3842/sigma.2007.105 2007
[33]

Albeverio and M

S. Albeverio and M. V. Altaisky, A remark on gauge invariance in wavelet-based quantum field theory (2010), arXiv:0901.2806 [hep-th]

work page arXiv 2010
[34]

M. V. Altaisky, Wavelets and renormalization group in quantum field theory problems, Physics of Atomic Nuclei 81, 786 (2018)

2018
[35]

M. V. Altaisky and N. E. Kaputkina, Continuous wavelet transform in quantum field theory, Phys. Rev. D88, 025015 (2013)

2013
[36]

M. V. Altaiskii and N. E. Kaputkina, On the wavelet decomposition in light cone variables, Russian Physics Journal55, 1177 (2013)

2013
[37]

M. V. Altaisky and N. E. Kaputkina, On quantization in light-cone variables compatible with wavelet transform, International Journal of Theoretical Physics55, 2805 (2016)

2016
[38]

W. N. Polyzou, Light-front quantum mechanics and quantum field theory (2020), arXiv:2002.04981 [hep-th]

work page arXiv 2020
[39]

Basak and R

M. Basak and R. Ratabole, Renormalization in a wavelet basis, Phys. Rev. D107, 036015 (2023)

2023
[40]

M.BasakandR.Ratabole,Hamiltonianflowequationsin the daubechies wavelet basis, Phys. Rev. D111, 096024 (2025)

2025
[41]

Basak, D

M. Basak, D. Chakraborty, N. Mathur, and R. Ratabole, Hamiltonian formulation of the(1 + 1)-dimensionalϕ 4 theory in a momentum-space daubechies wavelet basis, Phys. Rev. D , (2026)

2026
[42]

W. N. Polyzou, Wavelet representation of light-front quantum field theory, Phys. Rev. D101, 096004 (2020)

2020
[43]

B. M. Kessler, G. L. Payne, and W. N. Polyzou, Scattering calculations with wavelets, Few-Body Systems 33, 1 (2003)

2003
[44]

Basak, Multiresolution analysis of quantum theories using daubechies wavelet basis (2025), arXiv:2512.18372 [hep-th]

M. Basak, Multiresolution analysis of quantum theories using daubechies wavelet basis (2025), arXiv:2512.18372 [hep-th]

work page arXiv 2025
[45]

Michlin, W

T. Michlin, W. N. Polyzou, and F. Bulut, Multiresolution decomposition of quantum field theories using wavelet bases, Phys. Rev. D95, 094501 (2017)

2017
[46]

F.Wegner,Flow-equationsforhamiltonians,Annalender physik506, 77 (1994)

1994
[47]

S. D. Głazek and K. G. Wilson, Renormalization of hamiltonians, Physical Review D48, 5863 (1993)

1993
[48]

S. D. Glazek and K. G. Wilson, Perturbative renormalization group for hamiltonians, Physical Review D49, 4214 (1994)

1994
[49]

B. M. Kessler, G. L. Payne, and W. N. Polyzou, Wavelet notes (2003), arXiv:nucl-th/0305025 [nucl-th]

work page arXiv 2003
[50]

Daubechies,Ten Lectures on Wavelets, CBMS-NSF Regional Conference Series in Applied Mathematics (Society for Industrial and Applied Mathematics, 1992)

I. Daubechies,Ten Lectures on Wavelets, CBMS-NSF Regional Conference Series in Applied Mathematics (Society for Industrial and Applied Mathematics, 1992)

1992
[51]

F.Wegner,Flow-equationsforhamiltonians,Annalender Physik506, 77 (1994)

1994
[52]

S. D. Głazek and R. J. Perry, Impact of bound states on similarity renormalization group transformations, Phys. Rev. D78, 045011 (2008)

2008