pith. sign in

arxiv: 2607.02078 · v1 · pith:LWDRG3QBnew · submitted 2026-07-02 · ⚛️ physics.ins-det · hep-ex· physics.data-an

WavePID: Low-energy flavor identification using single-PMT time series in IceCube

R. Abbasi , M. Ackermann , J. Adams , J. A. Aguilar , M. Ahlers , J.M. Alameddine , S. Ali , N. M. Amin
show 411 more authors
K. Andeen C. Arg\"uelles S. Athanasiadou S. N. Axani R. Babu X. Bai A. Balagopal V. S. W. Barwick V. Basu R. Bay J. J. Beatty J. Becker Tjus P. Behrens J. Beise C. Bellenghi S. Benkel S. BenZvi D. Berley E. Bernardini D. Z. Besson E. Blaufuss L. Bloom S. Blot F. Bontempo J. Y. Book Motzkin C. Boscolo Meneguolo S. B\"oser O. Botner J. B\"ottcher J. Braun B. Brinson Z. Brisson-Tsavoussis L. Brusa R. T. Burley D. Butterfield K. Carloni J. Carpio N. Chau Y. C. Chen Z. Chen D. Chirkin S. Choi A. Chubarov B. A. Clark G. H. Collin D. A. Coloma Borja A. Connolly J. M. Conrad D. F. Cowen C. De Clercq J. J. DeLaunay D. Delgado T. Delmeulle S. Deng P. Desiati K. D. de Vries G. de Wasseige T. DeYoung J. C. D\'iaz-V\'elez S. DiKerby T. Ding M. Dittmer A. Domi L. Draper L. Dueser D. Durnford K. Dutta M. A. DuVernois T. Ehrhardt L. Eidenschink A. Eimer C. Eldridge P. Eller E. Ellinger D. Els\"asser R. Engel H. Erpenbeck W. Esmail S. Eulig J. Evans P. A. Evenson K. L. Fan K. Fang K. Farrag A. Fattorini A. R. Fazely A. Fedynitch N. Feigl C. Finley D. Fox A. Franckowiak S. Fukami P. F\"urst J. Gallagher E. Ganster A. Garcia M. Garcia E. Genton L. Gerhardt A. Ghadimi C. Glaser T. Gl\"usenkamp J. G. Gonzalez S. Goswami A. Granados D. Grant S. J. Gray S. Griffin S. Griswold K. M. Groth D. Guevel C. G\"unther P. Gutjahr C. Ha A. Hallgren L. Halve F. Halzen L. Hamacher M. Handt K. Hanson J. Hardin A. A. Harnisch P. Hatch A. Haungs J. H\"au{\ss}ler K. Helbing J. Hellrung B. Henke L. Hennig F. Henningsen L. Heuermann R. Hewett N. Heyer S. Hickford A. Hidvegi C. Hill G. C. Hill R. Hmaid K. D. Hoffman A. Hollnagel D. Hooper S. Hori K. Hoshina M. Hostert W. Hou M. Hrywniak T. Huber K. Hultqvist K. Hymon A. Ishihara W. Iwakiri M. Jacquart S. Jain O. Janik M. Jansson M. Jin N. Kamp D. Kang W. Kang A. Kappes L. Kardum T. Karg A. Karle A. Katil M. Kauer J. L. Kelley M. Khanal A. Khatee Zathul A. Kheirandish T. Kim H. Kimku F. Kirchner J. Kiryluk C. Klein S. R. Klein Y. Kobayashi S. Koch A. Kochocki R. Koirala H. Kolanoski T. Kontrimas L. K\"opke C. Kopper D. J. Koskinen P. Koundal M. Kowalski T. Kozynets A. Kravka N. Krieger T. Krishnan K. Kruiswijk E. Krupczak E. Kun N. Kurahashi C. Lagunas Gualda L. Lallement Arnaud M. J. Larson F. Lauber J. P. Lazar K. Leonard DeHolton A. Leszczy\'nska C. Li J. Liao C. Lin Q. R. Liu Y. T. Liu M. Liubarska C. Love L. Lu F. Lucarelli W. Luszczak Y. Lyu M. Macdonald E. Magnus Y. Makino E. Manao S. Mancina A. Mand I. C. Mari\c{s} S. Marka Z. Marka L. Marten I. Martinez-Soler R. Maruyama J. Mauro F. Mayhew F. McNally K. Meagher A. Medina M. Meier Y. Merckx L. Merten S. Minji J. Mitchell L. Molchany S. Mondal T. Montaruli R. W. Moore Y. Morii A. Mosbrugger D. Mousadi E. Moyaux T. Mukherjee M. Nakos U. Naumann R. Neshat L. Neste M. Neumann H. Niederhausen M. U. Nisa K. Noda A. Noell A. Novikov A. Obertacke V. O'Dell A. Olivas R. Orsoe J. Osborn E. O'Sullivan B. Owens V. Palusova H. Pandya A. Parenti C. Parisel N. Park V. Parrish E. N. Paudel L. Paul C. P\'erez de los Heros T. Pernice T. C. Petersen J. Peterson S. Pick M. Plum A. Pont\'en V. Poojyam B. Pries R. Procter-Murphy G. T. Przybylski L. Pyras C. Raab J. Rack-Helleis N. Rad M. Ravn K. Rawlins Z. Rechav A. Rehman I. Reistroffer E. Resconi C. D. Rho W. Rhode L. Ricca B. Riedel A. Rifaie E. J. Roberts S. Rodan M. Rongen A. Rosted C. Rott T. Ruhe L. Ruohan D. Ryckbosch J. Saffer D. Salazar-Gallegos P. Sampathkumar A. Sandrock G. Sanger-Johnson M. Santander S. Sarkar M. Scarnera M. Schaufel H. Schieler S. Schindler L. Schlickmann B. Schl\"uter F. Schl\"uter N. Schmeisser T. Schmidt A. Scholz F. G. Schr\"oder S. Schwirn S. Sclafani D. Seckel L. Seen M. Seikh S. Seunarine P. A. Sevle Myhr R. Shah S. Shah S. Shefali N. Shimizu B. Skrzypek R. Snihur J. Soedingrekso D. Soldin P. Soldin G. Sommani D. Song C. Spannfellner G. M. Spiczak C. Spiering J. Stachurska M. Stamatikos T. Stanev T. Stezelberger T. St\"urwald T. Stuttard G. W. Sullivan I. Taboada S. Ter-Antonyan A. Terliuk A. Thakuri M. Thiesmeyer W. G. Thompson J. Thwaites S. Tilav K. Tollefson J. A. Torres S. Toscano D. Tosi K. Upshaw A. Vaidyanathan N. Valtonen-Mattila J. Valverde J. Vandenbroucke T. Van Eeden N. van Eijndhoven L. Van Rootselaar J. van Santen J. Vara F. Varsi M. Velazquez M. Venugopal M. Vereecken S. Vergara Carrasco S. Verpoest D. Veske A. Vijai J. Villarreal C. Walck A. Wang E. H. S. Warrick C. Weaver P. Weigel A. Weindl J. Weldert A. Y. Wen C. Wendt J. Werthebach M. Weyrauch N. Whitehorn C. H. Wiebusch D. R. Williams L. Witthaus G. Wrede X. W. Xu J. P. Yanez Y. Yao E. Yildizci S. Yoshida F. Yu S. Yu T. Yuan S. Yun-C\'rcamo A. Zander Jurowitzki A. Zegarelli S. Zhang Z. Zhang P. Zhelnin P. Zilberman C. Zilleruelo Ca\~nas
This is my paper

Pith reviewed 2026-07-03 02:57 UTC · model grok-4.3

classification ⚛️ physics.ins-det hep-exphysics.data-an
keywords IceCubeneutrino flavor identificationpulse timingCherenkov radiationlow-energy neutrinosgraph neural networkparticle classificationtemplate likelihood
0
0 comments X

The pith

Per-module pulse timing carries flavor-identification information complementary to morphology-based classifiers in IceCube.

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

The paper introduces WavePID, a template-based log-likelihood-ratio classifier that uses three nanosecond-scale timing observables from single detector modules to identify neutrino flavors at low energies. Evaluated on a cascade-enriched sample chosen by a graph neural network, WavePID raises both cascade purity and overall classification performance. A sympathetic reader would care because sparse photon detection makes flavor classification difficult in the 5-100 GeV range needed for oscillation measurements and beyond-Standard-Model searches. Geant4 simulations tie the timing signal to differences in Cherenkov emission geometry between muon tracks and electromagnetic showers.

Core claim

WavePID is a template-based log-likelihood-ratio classifier that exploits three timing observables on individual detector modules: distance to the reconstructed vertex, early-charge fraction, and module-to-module time difference. When evaluated on a cascade-enriched sample selected by a state-of-the-art graph neural network, WavePID improves both cascade purity and classification performance over the neural network alone. This demonstrates that per-module pulse timing carries flavor-identification information complementary to morphology-based classifiers. Geant4 simulations associate this signal with differences in Cherenkov emission geometry between muon tracks and electromagnetic showers.

What carries the argument

WavePID, a template-based log-likelihood-ratio classifier that uses per-module timing observables (distance to reconstructed vertex, early-charge fraction, module-to-module time difference) to separate neutrino flavors.

If this is right

  • WavePID improves cascade purity and classification performance beyond the graph neural network alone.
  • Per-module pulse timing provides flavor-identification information complementary to morphology-based methods.
  • Nanosecond-scale pulse timing becomes a usable observable for low-energy neutrino reconstruction.
  • Detector designs with improved per-module timing resolution can exploit the same timing observables.

Where Pith is reading between the lines

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

  • The timing-based approach could be tested on data from other single-PMT neutrino detectors to check if Cherenkov geometry differences remain useful.
  • Combining WavePID-style timing with additional observables might further raise performance in oscillation or new-physics analyses.
  • The method could be extended to other energy ranges if the underlying emission-geometry differences persist in simulations.

Load-bearing premise

The timing observables supply information independent of the graph neural network morphology features when evaluated on a GNN-selected cascade-enriched sample.

What would settle it

A test showing no improvement in cascade purity when WavePID is applied to the GNN-selected sample, or Geant4 simulations revealing identical timing distributions for muon-track and electromagnetic-shower events, would falsify the central claim.

read the original abstract

The IceCube Neutrino Observatory, a cubic-kilometer detector at the South Pole, identifies neutrino flavor through event morphology. Sparse photon detection makes this classification particularly challenging in the 5--100~GeV regime, the energy range relevant for oscillation measurements and searches for physics beyond the Standard Model. We introduce WavePID, a template-based log-likelihood-ratio classifier that exploits nanosecond-scale timing on individual detector modules through three observables: the distance to the reconstructed vertex, the early-charge fraction, and the module-to-module time difference. Evaluated on a cascade-enriched sample selected by a state-of-the-art graph neural network, WavePID improves both cascade purity and classification performance over the neural network alone. This demonstrates that per-module pulse timing carries flavor-identification information complementary to morphology-based classifiers, opening a new physics-motivated observable for low-energy neutrino reconstruction. Geant4 simulations associate this signal with differences in Cherenkov emission geometry between muon tracks and electromagnetic showers. These results motivate exploiting nanosecond-scale pulse timing in future low-energy classifiers and in detector designs with improved per-module timing in next-generation neutrino telescopes.

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.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces WavePID, a template-based log-likelihood-ratio classifier that uses three single-PMT timing observables (distance to reconstructed vertex, early-charge fraction, module-to-module time difference) to improve neutrino flavor identification at 5-100 GeV in IceCube. Evaluated on a cascade-enriched sample pre-selected by a state-of-the-art graph neural network, it reports gains in cascade purity and overall classification performance relative to the GNN alone, with Geant4 simulations linking the signal to Cherenkov geometry differences between muon tracks and electromagnetic showers.

Significance. If the timing observables supply statistically independent information, the result would establish a new, physically motivated observable for low-energy flavor tagging that is complementary to morphology-based methods. This could directly benefit oscillation analyses and BSM searches in IceCube and motivate per-module timing upgrades in next-generation detectors. The explicit Geant4 connection to emission geometry is a positive feature.

major comments (2)
  1. [Abstract and evaluation section] The central claim that the three timing observables carry flavor-identification information complementary to the GNN (Abstract; evaluation section) rests on an unverified assumption of statistical independence. Because the GNN operates on per-module hit times as node features, an ablation study or correlation analysis between the derived observables and GNN internal representations is required to rule out redundancy; without it the reported improvement on the GNN-selected sample could arise from the LLR functional form or post-selection effects rather than new information.
  2. [Abstract] No quantitative performance metrics (e.g., purity values, ROC-AUC deltas, error bars, or template-construction details) are supplied in the Abstract or referenced in the evaluation, preventing assessment of whether the claimed improvement is statistically significant or practically meaningful.
minor comments (2)
  1. [Methods section] Clarify the exact definition and binning of the three timing observables and how the templates are constructed from simulation.
  2. [Evaluation section] Add a brief statement on the size of the cascade-enriched test sample and any cross-validation procedure used.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed review. We respond point by point to the major comments below, indicating planned revisions where appropriate.

read point-by-point responses

  1. Referee: [Abstract and evaluation section] The central claim that the three timing observables carry flavor-identification information complementary to the GNN (Abstract; evaluation section) rests on an unverified assumption of statistical independence. Because the GNN operates on per-module hit times as node features, an ablation study or correlation analysis between the derived observables and GNN internal representations is required to rule out redundancy; without it the reported improvement on the GNN-selected sample could arise from the LLR functional form or post-selection effects rather than new information.

    Authors: We agree that explicitly demonstrating statistical independence would strengthen the complementarity claim. The WavePID observables are constructed as template-based LLR inputs from single-PMT time series, distinct from the GNN node features, yet we acknowledge the possibility of overlap. In the revised manuscript we will add both a correlation analysis between the three WavePID observables and the GNN internal representations and an ablation study that removes timing-derived information from the GNN, allowing quantitative assessment of redundancy versus new information. revision: yes

  2. Referee: [Abstract] No quantitative performance metrics (e.g., purity values, ROC-AUC deltas, error bars, or template-construction details) are supplied in the Abstract or referenced in the evaluation, preventing assessment of whether the claimed improvement is statistically significant or practically meaningful.

    Authors: The evaluation section reports quantitative gains in cascade purity and classification performance together with error estimates and template-construction details. To improve accessibility we will revise the Abstract to include the principal numerical results (purity deltas, ROC-AUC changes) and add explicit cross-references to the relevant evaluation figures and tables. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical improvement on independent observables

full rationale

The paper's central result is an empirical demonstration that three explicitly defined timing observables (distance to vertex, early-charge fraction, module-to-module time difference) yield measurable gains in purity and classification when applied to a GNN-preselected cascade sample. No equations, parameter fits, or definitions are shown that reduce the reported improvement to the inputs by construction. The observables are derived directly from per-PMT time series data and are presented as physically motivated by Cherenkov geometry differences, with the performance gain serving as the test of complementarity rather than a self-referential claim. No load-bearing self-citations or uniqueness theorems appear in the provided text. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on Geant4 accurately modeling Cherenkov geometry differences and on the timing observables being complementary to GNN features; no explicit free parameters are listed in the abstract.

axioms (1)
  • domain assumption Geant4 simulations correctly associate observed timing differences with Cherenkov emission geometry variations between muon tracks and electromagnetic showers
    Invoked in the abstract to explain the physical origin of the timing signal.

pith-pipeline@v0.9.1-grok · 8018 in / 1094 out tokens · 38189 ms · 2026-07-03T02:57:24.397868+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

31 extracted references · 17 canonical work pages · 4 internal anchors

  1. [1]

    2025 , howpublished =

    2025

  2. [2]

    Aartsen, M. G. and others , collaboration =. Calibration and characterization of the. Nucl. Instrum. Meth. A , volume =. 2010 , eprint =

    2010

  3. [3]

    Aartsen, M. G. and others , collaboration =. Refining the IceCube detector geometry using muon and LED calibration data. PoS. doi:10.22323/1.444.0988

    work page doi:10.22323/1.444.0988

  4. [4]

    Calculation of the Cherenkov light yield from electromagnetic cascades in ice with Geant4 , journal =

    Leif Rädel and Christopher Wiebusch , keywords =. Calculation of the Cherenkov light yield from electromagnetic cascades in ice with Geant4 , journal =. 2013 , issn =. doi:https://doi.org/10.1016/j.astropartphys.2013.01.015 , url =

    work page doi:10.1016/j.astropartphys.2013.01.015 2013

  5. [5]

    Groom, D. E. and Klein, S. R. , title =. The European Physical Journal C - Particles and Fields , volume =. 2000 , month = mar, doi =

    2000

  6. [6]

    GROOM and NIKOLAI V

    DONALD E. GROOM and NIKOLAI V. MOKHOV and SERGEI I. STRIGANOV , abstract =. MUON STOPPING POWER AND RANGE TABLES 10 MeV–100 TeV , journal =. 2001 , issn =. doi:https://doi.org/10.1006/adnd.2001.0861 , url =

    work page doi:10.1006/adnd.2001.0861 2001

  7. [7]

    Atomic and Nuclear Properties of Water (Ice)

    Particle Data Group. Atomic and Nuclear Properties of Water (Ice). 2020

    2020

  8. [8]

    2019 , url =

    Large Photocathode Area Photomultiplier Tubes , author =. 2019 , url =

    2019

  9. [9]

    Axani , title =

    Spencer N. Axani , title =

  10. [10]

    2024 , eprint=

    Learning Efficient Representations of Neutrino Telescope Events , author=. 2024 , eprint=

    2024

  11. [11]

    Aartsen, M. G. and others , collaboration =. The. JINST , volume =. 2017 , doi =. 1612.05093 , archivePrefix =

    work page arXiv 2017

  12. [12]

    Aartsen, M. G. and others , collaboration =. Measurement of. Nucl. Instrum. Meth. A , volume =. 2013 , doi =. 1301.5361 , archivePrefix =

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  13. [13]

    and others , collaboration =

    Abbasi, R. and others , collaboration =. Low energy event reconstruction in. Eur. Phys. J. C , volume =. 2022 , doi =. 2203.02303 , archivePrefix =

    work page arXiv 2022

  14. [14]

    Abbasiet al.(IceCube), Measurement of atmospheric neutrino mixing with improved IceCube DeepCore cali- bration and data processing, Phys

    Abbasi, R. and others , collaboration =. Measurement of atmospheric neutrino mixing with improved. Phys. Rev. D , volume =. 2023 , doi =. 2304.12236 , archivePrefix =

    work page arXiv 2023

  15. [15]

    and others , collaboration =

    Abbasi, R. and others , collaboration =. Development of an analysis to probe the neutrino mass ordering with atmospheric neutrinos using three years of. Eur. Phys. J. C , volume =. 2020 , doi =. 1902.07771 , archivePrefix =

    work page arXiv 2020

  16. [16]

    and others , collaboration =

    Abbasi, R. and others , collaboration =. Search for a light sterile neutrino with 7.5 years of. Phys. Rev. D , volume =. 2024 , doi =. 2407.01314 , archivePrefix =

    work page arXiv 2024

  17. [17]

    and others , collaboration =

    Abbasi, R. and others , collaboration =. Graph neural networks for low-energy event classification and reconstruction in. JINST , volume =. 2022 , doi =. 2209.03042 , archivePrefix =

    work page arXiv 2022

  18. [18]

    and others , title =

    Samani, S. and others , title =. JINST , volume =. 2020 , doi =. 1912.03901 , archivePrefix =

    work page arXiv 2020

  19. [19]

    and others , title =

    Agostinelli, S. and others , title =. Nucl. Instrum. Meth. A , volume =. 2003 , doi =

    2003

  20. [20]

    2024 , month = nov, note =

    Kozynets, Tetiana , title =. 2024 , month = nov, note =

    2024

  21. [21]

    The Design and Performance of IceCube DeepCore

    Abbasi, R. and others. The Design and Performance of IceCube DeepCore. Astropart. Phys. 2012. doi:10.1016/j.astropartphys.2012.01.004. arXiv:1109.6096

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1016/j.astropartphys.2012.01.004 2012

  22. [22]

    Aartsen, M. G. and others. Measurement of Atmospheric Tau Neutrino Appearance with IceCube DeepCore. Phys. Rev. D. 2019. doi:10.1103/PhysRevD.99.032007. arXiv:1901.05366

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1103/physrevd.99.032007 2019

  23. [23]

    and others

    Abbasi, R. and others. Measurement of Atmospheric Neutrino Oscillation Parameters Using Convolutional Neural Networks with 9.3 Years of Data in IceCube DeepCore. Phys. Rev. Lett. 2025. doi:10.1103/PhysRevLett.134.091801. arXiv:2405.02163

    work page doi:10.1103/physrevlett.134.091801 2025

  24. [24]

    The GENIE Neutrino Monte Carlo Generator

    Andreopoulos, C. and others. The GENIE Neutrino Monte Carlo Generator. Nucl. Instrum. Meth. A. 2010. doi:10.1016/j.nima.2009.12.009. arXiv:0905.2517

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1016/j.nima.2009.12.009 2010

  25. [25]

    JINST , volume =

    The. JINST , volume =. 2022 , eprint =

    2022

  26. [26]

    Determining the neutrino mass ordering and oscillation parameters with. Eur. Phys. J. C , volume =. 2022 , eprint =

    2022

  27. [27]

    J. Phys. G , volume =. 2021 , eprint =

    2021

  28. [28]

    and others , collaboration =

    Agostini, M. and others , collaboration =. The. Nature Astron. , volume =. 2020 , eprint =

    2020

  29. [29]

    Ye, Z. P. and others , journal =. A multi-cubic-kilometre neutrino telescope in the western. 2023 , eprint =

    2023

  30. [30]

    Huang, Tian-Qi and Cao, Zhen and Chen, Mingjun and Liu, Jiali and Wang, Zike and You, Xiaohao and Qi, Ying , title = ". PoS. doi:10.22323/1.444.1080

    work page doi:10.22323/1.444.1080

  31. [31]

    Design and performance of the multi-PMT optical module for IceCube Upgrade. PoS. doi:10.22323/1.395.1070. arXiv:2107.11383

    work page doi:10.22323/1.395.1070