publications

2025

1. Preprint

   Fast and accurate parameter estimation of high-redshift sources with the Einstein Telescope

   Filippo Santoliquido, Jacopo Tissino, Ulyana Dupletsa, and 8 more authors

   Apr 2025

   arXiv Bib

   @article{Santoliquido:2025lot,
     author = {Santoliquido, Filippo and Tissino, Jacopo and Dupletsa, Ulyana and Branchesi, Marica and Harms, Jan and Sedda, Manuel Arca and Dax, Maximilian and Kofler, Annalena and Green, Stephen R and Gupte, Nihar and others},
     title = {{Fast and accurate parameter estimation of high-redshift sources with the Einstein Telescope}},
     eprint = {2504.21087},
     archiveprefix = {arXiv},
     primaryclass = {astro-ph.HE},
     month = apr,
     year = {2025},
   }

2. Preprint

   Reparameterized LLM Training via Orthogonal Equivalence Transformation

   Zeju Qiu, Simon Buchholz, Tim Z Xiao, and 3 more authors

   Jun 2025

   arXiv Bib

   @article{qiu2025reparameterized,
     title = {Reparameterized LLM Training via Orthogonal Equivalence Transformation},
     author = {Qiu, Zeju and Buchholz, Simon and Xiao, Tim Z and Dax, Maximilian and Sch{\"o}lkopf, Bernhard and Liu, Weiyang},
     eprint = {2506.08001},
     archiveprefix = {arXiv},
     year = {2025},
     month = jun,
   }

3. Nature

   

   Real-time inference for binary neutron star mergers using machine learning

   Maximilian Dax, Stephen R. Green, Jonathan Gair, and 7 more authors

   Nature, Mar 2025

   DOI arXiv Bib HTML PRESS PDF

   

   @article{Dax:2024mcn,
     author = {Dax, Maximilian and Green, Stephen R. and Gair, Jonathan and Gupte, Nihar and P{\"u}rrer, Michael and Raymond, Vivien and Wildberger, Jonas and Macke, Jakob H. and Buonanno, Alessandra and Sch{\"o}lkopf, Bernhard},
     title = {{Real-time inference for binary neutron star mergers using machine learning}},
     eprint = {2407.09602},
     archiveprefix = {arXiv},
     primaryclass = {gr-qc},
     reportnumber = {LIGO-P2400294},
     doi = {10.1038/s41586-025-08593-z},
     journal = {Nature},
     volume = {639},
     number = {8053},
     pages = {49--53},
     year = {2025},
     month = mar,
     url = {https://www.nature.com/articles/s41586-025-08593-z},
     dimensions = {true},
   }

4. Workshop Paper

   Synthesizing 3D Abstractions by Inverting Procedural Buildings with Transformers

   Maximilian Dax, Jordi Serrano Berbel, Jan Stria, and 2 more authors

   In CVPR workshop on Synthetic Data for Computer Vision Workshop, Jun 2025

   arXiv Bib

   @inproceedings{dax2025synthesizing,
     title = {Synthesizing 3D Abstractions by Inverting Procedural Buildings with Transformers},
     author = {Dax, Maximilian and Berbel, Jordi Serrano and Stria, Jan and Guibas, Leonidas and Bergmann, Urs M},
     booktitle = {CVPR workshop on Synthetic Data for Computer Vision Workshop},
     eprint = {2501.17044},
     archiveprefix = {arXiv},
     year = {2025},
     month = jun,
   }

5. A&A

   Flow matching for atmospheric retrieval of exoplanets: Where reliability meets adaptive noise levels

   Timothy D Gebhard, Jonas Wildberger, Maximilian Dax, and 4 more authors

   Astronomy & Astrophysics, Jun 2025

   Bib

   @article{gebhard2024flow,
     title = {Flow matching for atmospheric retrieval of exoplanets: Where reliability meets adaptive noise levels},
     author = {Gebhard, Timothy D and Wildberger, Jonas and Dax, Maximilian and Kofler, Annalena and Angerhausen, Daniel and Quanz, Sascha P and Sch{\"o}lkopf, Bernhard},
     journal = {Astronomy \& Astrophysics},
     volume = {693},
     pages = {A42},
     year = {2025},
     publisher = {EDP Sciences},
   }

6. Science Advances

   Fast and reliable probabilistic reflectometry inversion with prior-amortized neural posterior estimation

   Vladimir Starostin, Maximilian Dax, Alexander Gerlach, and 3 more authors

   Science Advances, Jun 2025

   Bib

   @article{starostin2025fast,
     title = {Fast and reliable probabilistic reflectometry inversion with prior-amortized neural posterior estimation},
     author = {Starostin, Vladimir and Dax, Maximilian and Gerlach, Alexander and Hinderhofer, Alexander and Tejero-Cantero, {\'A}lvaro and Schreiber, Frank},
     journal = {Science Advances},
     volume = {11},
     number = {11},
     pages = {eadr9668},
     year = {2025},
     publisher = {American Association for the Advancement of Science},
   }

2024

1. Preprint

   Evidence for eccentricity in the population of binary black holes observed by LIGO-Virgo-KAGRA

   Nihar Gupte, Antoni Ramos-Buades, Alessandra Buonanno, and 9 more authors

   Apr 2024

   arXiv Bib

   @misc{Gupte:2024jfe,
     author = {Gupte, Nihar and Ramos-Buades, Antoni and Buonanno, Alessandra and Gair, Jonathan and Miller, M Coleman and Dax, Maximilian and Green, Stephen R and P{\"u}rrer, Michael and Wildberger, Jonas and Macke, Jakob and Romero-Shaw, Isobel M. and Sch{\"o}lkopf, Bernhard},
     title = {{Evidence for eccentricity in the population of binary black holes observed by LIGO-Virgo-KAGRA}},
     eprint = {2404.14286},
     archiveprefix = {arXiv},
     primaryclass = {gr-qc},
     month = apr,
     year = {2024},
   }

2. Workshop Paper

   Inferring Atmospheric Properties of Exoplanets with Flow Matching and Neural Importance Sampling

   Timothy D Gebhard, Jonas Wildberger, Maximilian Dax, and 3 more authors

   In AAAI Workshop on AI to Accelerate Science and Engineering (spotlight), Apr 2024

   Bib

   @inproceedings{gebhard2023inferring,
     title = {Inferring Atmospheric Properties of Exoplanets with Flow Matching and Neural Importance Sampling},
     author = {Gebhard, Timothy D and Wildberger, Jonas and Dax, Maximilian and Angerhausen, Daniel and Quanz, Sascha P and Sch{\"o}lkopf, Bernhard},
     booktitle = {AAAI Workshop on AI to Accelerate Science and Engineering (spotlight)},
     year = {2024},
     keywords = {W},
   }

2023

1. PRL

   

   Neural Importance Sampling for Rapid and Reliable Gravitational-Wave Inference

   Maximilian Dax, Stephen R. Green, Jonathan Gair, and 5 more authors

   Phys. Rev. Lett., Apr 2023

   DOI arXiv Bib HTML PRESS PDF

   

   @article{Dax:2022pxd,
     author = {Dax, Maximilian and Green, Stephen R. and Gair, Jonathan and P{\"u}rrer, Michael and Wildberger, Jonas and Macke, Jakob H. and Buonanno, Alessandra and Sch{\"o}lkopf, Bernhard},
     title = {{Neural Importance Sampling for Rapid and Reliable Gravitational-Wave Inference}},
     eprint = {2210.05686},
     archiveprefix = {arXiv},
     primaryclass = {gr-qc},
     reportnumber = {LIGO-P2200297},
     doi = {10.1103/PhysRevLett.130.171403},
     journal = {Phys. Rev. Lett.},
     volume = {130},
     number = {17},
     pages = {171403},
     year = {2023},
     month = apr,
     url = {https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.130.171403},
     dimensions = {true},
   }

2. NeurIPS

   Flow Matching for Scalable Simulation-Based Inference

   Jonas Wildberger^*, Maximilian Dax^*, Simon Buchholz^*, and 3 more authors

   NeurIPS 2023, Dec 2023

   arXiv Bib

   @article{dax2023flow,
     title = {Flow Matching for Scalable Simulation-Based Inference},
     author = {Wildberger, Jonas and Dax, Maximilian and Buchholz, Simon and Green, Stephen R and Macke, Jakob H and Sch{\"o}lkopf, Bernhard},
     journal = {NeurIPS 2023},
     eprint = {2305.17161},
     archiveprefix = {arXiv},
     year = {2023},
     month = dec,
   }

3. PRD

   Adapting to noise distribution shifts in flow-based gravitational-wave inference

   Jonas Wildberger, Maximilian Dax, Stephen R. Green, and 5 more authors

   Phys. Rev. D, Dec 2023

   DOI arXiv Bib

   @article{Wildberger:2022agw,
     author = {Wildberger, Jonas and Dax, Maximilian and Green, Stephen R. and Gair, Jonathan and P\"urrer, Michael and Macke, Jakob H. and Buonanno, Alessandra and Sch\"olkopf, Bernhard},
     title = {{Adapting to noise distribution shifts in flow-based gravitational-wave inference}},
     eprint = {2211.08801},
     archiveprefix = {arXiv},
     primaryclass = {gr-qc},
     doi = {10.1103/PhysRevD.107.084046},
     journal = {Phys. Rev. D},
     volume = {107},
     number = {8},
     pages = {084046},
     year = {2023},
   }

4. Workshop Paper

   Flow Matching for Scalable Simulation-Based Inference

   Jonas Wildberger^*, Maximilian Dax^*, Simon Buchholz^*, and 3 more authors

   In ICML 2023 Workshop on Machine Learning for Astrophysics (spotlight), Jul 2023

   Bib

   @inproceedings{wildberger2023flow,
     title = {Flow Matching for Scalable Simulation-Based Inference},
     author = {Wildberger, Jonas and Dax, Maximilian and Buchholz, Simon and Green, Stephen R and Macke, Jakob H and Sch{\"o}lkopf, Bernhard},
     booktitle = {ICML 2023 Workshop on Machine Learning for Astrophysics (spotlight)},
     year = {2023},
     month = jul,
     keywords = {W},
   }

5. Workshop Paper

   Flow Matching for Scalable Simulation-Based Inference

   Jonas Bernhard Wildberger^*, Maximilian Dax^*, Simon Buchholz^*, and 3 more authors

   In ICML 2023 Workshop on Structured Probabilistic Inference & Generative Modeling, Jul 2023

   Bib

   @inproceedings{wildberger2023flow2,
     title = {Flow Matching for Scalable Simulation-Based Inference},
     author = {Wildberger, Jonas Bernhard and Dax, Maximilian and Buchholz, Simon and Green, Stephen R and Macke, Jakob H. and Sch{\"o}lkopf, Bernhard},
     booktitle = {ICML 2023 Workshop on Structured Probabilistic Inference {\&} Generative Modeling},
     year = {2023},
     month = jul,
     keywords = {W},
     url = {https://openreview.net/forum?id=LdGjxxjfh8},
   }

2022

1. ICLR

   Group equivariant neural posterior estimation

   Maximilian Dax, Stephen R. Green, Jonathan Gair, and 3 more authors

   ICLR 2022, Jul 2022

   arXiv Bib

   @article{Dax:2021myb,
     author = {Dax, Maximilian and Green, Stephen R. and Gair, Jonathan and Deistler, Michael and Sch\"olkopf, Bernhard and Macke, Jakob H.},
     title = {{Group equivariant neural posterior estimation}},
     journal = {ICLR 2022},
     year = {2022},
     eprint = {2111.13139},
     archiveprefix = {arXiv},
     primaryclass = {cs.LG},
   }

2. Workshop Paper

   Addressing out-of-distribution data for flow-based gravitational wave inference

   Maximilian Dax^*, Stephen R Green^*, Jonas Wildberger^*, and 5 more authors

   In Machine Learning and the Physical Sciences Workshop at NeurIPS 2022, Dec 2022

   Bib

   @inproceedings{daxaddressing,
     title = {Addressing out-of-distribution data for flow-based gravitational wave inference},
     author = {Dax, Maximilian and Green, Stephen R and Wildberger, Jonas and Gair, Jonathan and P{\"u}rrer, Michael and Macke, Jakob H and Buonanno, Alessandra and Sch{\"o}lkopf, Bernhard},
     booktitle = {Machine Learning and the Physical Sciences Workshop at NeurIPS 2022},
     year = {2022},
     month = dec,
     keywords = {W},
   }

2021

1. PRL

   

   Real-Time Gravitational Wave Science with Neural Posterior Estimation

   Maximilian Dax, Stephen R. Green, Jonathan Gair, and 3 more authors

   Phys. Rev. Lett., Dec 2021

   DOI arXiv Bib HTML PRESS PDF

   

   @article{Dax:2021tsq,
     author = {Dax, Maximilian and Green, Stephen R. and Gair, Jonathan and Macke, Jakob H. and Buonanno, Alessandra and Sch\"olkopf, Bernhard},
     title = {{Real-Time Gravitational Wave Science with Neural Posterior Estimation}},
     eprint = {2106.12594},
     archiveprefix = {arXiv},
     primaryclass = {gr-qc},
     reportnumber = {LIGO-P2100223},
     doi = {10.1103/PhysRevLett.127.241103},
     journal = {Phys. Rev. Lett.},
     volume = {127},
     number = {24},
     pages = {241103},
     year = {2021},
     url = {https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.127.241103},
     dimensions = {true},
   }

2. EPJC

   Dispersive analysis of the Primakoff reaction γK →K \pi

   Maximilian Dax, Dominik Stamen, and Bastian Kubis

   Eur. Phys. J. C, Dec 2021

   DOI arXiv Bib

   @article{Dax:2020dzg,
     author = {Dax, Maximilian and Stamen, Dominik and Kubis, Bastian},
     title = {{Dispersive analysis of the Primakoff reaction $\gamma K \rightarrow K \pi $}},
     eprint = {2012.04655},
     archiveprefix = {arXiv},
     primaryclass = {hep-ph},
     doi = {10.1140/epjc/s10052-021-08951-x},
     journal = {Eur. Phys. J. C},
     volume = {81},
     number = {3},
     pages = {221},
     year = {2021},
   }

3. AAAI

   Explicitly modeled attention maps for image classification

   Andong Tan, Duc Tam Nguyen, Maximilian Dax, and 2 more authors

   AAAI 2021, Dec 2021

   arXiv Bib

   @article{tan2021explicitly,
     title = {Explicitly modeled attention maps for image classification},
     author = {Tan, Andong and Nguyen, Duc Tam and Dax, Maximilian and Nie{\ss}ner, Matthias and Brox, Thomas},
     journal = {AAAI 2021},
     volume = {35},
     number = {11},
     pages = {9799--9807},
     year = {2021},
     eprint = {2006.07872},
     archiveprefix = {arXiv},
     primaryclass = {cs.CV},
   }

4. Workshop Paper

   Amortized Bayesian inference of gravitational waves with normalizing flows

   Maximilian Dax, Stephen R Green, Jonathan Gair, and 3 more authors

   In Machine Learning and the Physical Sciences Workshop at NeurIPS 2021 (contributed talk), Dec 2021

   Bib

   @inproceedings{daxamortized,
     title = {Amortized Bayesian inference of gravitational waves with normalizing flows},
     author = {Dax, Maximilian and Green, Stephen R and Gair, Jonathan and Macke, Jakob H and Buonanno, Alessandra and Sch{\"o}lkopf, Bernhard},
     booktitle = {Machine Learning and the Physical Sciences Workshop at NeurIPS 2021 (contributed talk)},
     year = {2021},
     month = dec,
     keywords = {W},
   }

2019

1. NeurIPS

   DeepUSPS: Deep Robust Unsupervised Saliency Prediction With Self-Supervision

   Tam Nguyen^*, Maximilian Dax^*, Chaithanya Kumar Mummadi, and 4 more authors

   NeurIPS 2019, Dec 2019

   arXiv Bib

   @article{nguyen2019deepusps,
     title = {{DeepUSPS: Deep Robust Unsupervised Saliency Prediction With Self-Supervision}},
     author = {Nguyen, Tam and Dax, Maximilian and Mummadi, Chaithanya Kumar and Ngo, Nhung and Nguyen, Thi Hoai Phuong and Lou, Zhongyu and Brox, Thomas},
     journal = {NeurIPS 2019},
     volume = {33},
     year = {2019},
     eprint = {1909.13055},
     archiveprefix = {arXiv},
     primaryclass = {cs.CV},
   }

2018

1. EPJC

   Quark-mass dependence in ω→3\pi decays

   Maximilian Dax, Tobias Isken, and Bastian Kubis

   Eur. Phys. J. C, Dec 2018

   DOI arXiv Bib

   @article{Dax:2018rvs,
     author = {Dax, Maximilian and Isken, Tobias and Kubis, Bastian},
     title = {{Quark-mass dependence in $\omega \rightarrow 3\pi $ decays}},
     eprint = {1808.08957},
     archiveprefix = {arXiv},
     primaryclass = {hep-ph},
     doi = {10.1140/epjc/s10052-018-6346-3},
     journal = {Eur. Phys. J. C},
     volume = {78},
     number = {10},
     pages = {859},
     year = {2018},
   }