[{"title":"Fast symbolic algorithms for mega-regular games under strong transition fairness","intvolume":"         2","acknowledgement":"A previous version of this paper has appeared in TACAS 2022. Authors ordered alphabetically. T. Banerjee was interning with MPI-SWS when this research was conducted. R. Majumdar and A.-K. Schmuck are partially supported by DFG project 389792660 TRR 248–CPEC. A.-K. Schmuck is additionally funded through DFG project (SCHM 3541/1-1). K. Mallik is supported by the ERC project ERC-2020-AdG 101020093.","year":"2023","language":[{"iso":"eng"}],"article_processing_charge":"Yes","publication":"TheoretiCS","ddc":["000"],"month":"02","date_published":"2023-02-24T00:00:00Z","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)","image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)"},"type":"journal_article","citation":{"chicago":"Banerjee, Tamajit, Rupak Majumdar, Kaushik Mallik, Anne-Kathrin Schmuck, and Sadegh Soudjani. “Fast Symbolic Algorithms for Mega-Regular Games under Strong Transition Fairness.” <i>TheoretiCS</i>. EPI Sciences, 2023. <a href=\"https://doi.org/10.46298/theoretics.23.4\">https://doi.org/10.46298/theoretics.23.4</a>.","ista":"Banerjee T, Majumdar R, Mallik K, Schmuck A-K, Soudjani S. 2023. Fast symbolic algorithms for mega-regular games under strong transition fairness. TheoretiCS. 2, 4.","short":"T. Banerjee, R. Majumdar, K. Mallik, A.-K. Schmuck, S. Soudjani, TheoretiCS 2 (2023).","apa":"Banerjee, T., Majumdar, R., Mallik, K., Schmuck, A.-K., &#38; Soudjani, S. (2023). Fast symbolic algorithms for mega-regular games under strong transition fairness. <i>TheoretiCS</i>. EPI Sciences. <a href=\"https://doi.org/10.46298/theoretics.23.4\">https://doi.org/10.46298/theoretics.23.4</a>","mla":"Banerjee, Tamajit, et al. “Fast Symbolic Algorithms for Mega-Regular Games under Strong Transition Fairness.” <i>TheoretiCS</i>, vol. 2, 4, EPI Sciences, 2023, doi:<a href=\"https://doi.org/10.46298/theoretics.23.4\">10.46298/theoretics.23.4</a>.","ieee":"T. Banerjee, R. Majumdar, K. Mallik, A.-K. Schmuck, and S. Soudjani, “Fast symbolic algorithms for mega-regular games under strong transition fairness,” <i>TheoretiCS</i>, vol. 2. EPI Sciences, 2023.","ama":"Banerjee T, Majumdar R, Mallik K, Schmuck A-K, Soudjani S. Fast symbolic algorithms for mega-regular games under strong transition fairness. <i>TheoretiCS</i>. 2023;2. doi:<a href=\"https://doi.org/10.46298/theoretics.23.4\">10.46298/theoretics.23.4</a>"},"date_updated":"2024-02-05T10:21:51Z","has_accepted_license":"1","article_number":"4","oa_version":"Published Version","file_date_updated":"2024-02-05T10:19:35Z","article_type":"original","doi":"10.46298/theoretics.23.4","ec_funded":1,"quality_controlled":"1","oa":1,"arxiv":1,"day":"24","department":[{"_id":"ToHe"}],"date_created":"2024-01-31T13:40:49Z","status":"public","file":[{"relation":"main_file","file_id":"14940","file_size":917076,"content_type":"application/pdf","date_updated":"2024-02-05T10:19:35Z","creator":"dernst","success":1,"checksum":"2972d531122a6f15727b396110fb3f5c","date_created":"2024-02-05T10:19:35Z","file_name":"2023_TheoretiCS_Banerjee.pdf","access_level":"open_access"}],"publisher":"EPI Sciences","publication_status":"published","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","abstract":[{"text":"We consider fixpoint algorithms for two-player games on graphs with $\\omega$-regular winning conditions, where the environment is constrained by a strong transition fairness assumption. Strong transition fairness is a widely occurring special case of strong fairness, which requires that any execution is strongly fair with respect to a specified set of live edges: whenever the\r\nsource vertex of a live edge is visited infinitely often along a play, the edge itself is traversed infinitely often along the play as well. We show that, surprisingly, strong transition fairness retains the algorithmic characteristics of the fixpoint algorithms for $\\omega$-regular games -- the new algorithms have the same alternation depth as the classical algorithms but invoke a new type of predecessor operator. For Rabin games with $k$ pairs, the complexity of the new algorithm is $O(n^{k+2}k!)$ symbolic steps, which is independent of the number of live edges in the strong transition fairness assumption. Further, we show that GR(1) specifications with strong transition fairness assumptions can be solved with a 3-nested fixpoint algorithm, same as the usual algorithm. In contrast, strong fairness necessarily requires increasing the alternation depth depending on the number of fairness assumptions. We get symbolic algorithms for (generalized) Rabin, parity and GR(1) objectives under strong transition fairness assumptions as well as a direct symbolic algorithm for qualitative winning in stochastic\r\n$\\omega$-regular games that runs in $O(n^{k+2}k!)$ symbolic steps, improving the state of the art. Finally, we have implemented a BDD-based synthesis engine based on our algorithm. We show on a set of synthetic and real benchmarks that our algorithm is scalable, parallelizable, and outperforms previous algorithms by orders of magnitude.","lang":"eng"}],"_id":"14920","project":[{"name":"Vigilant Algorithmic Monitoring of Software","_id":"62781420-2b32-11ec-9570-8d9b63373d4d","grant_number":"101020093","call_identifier":"H2020"}],"publication_identifier":{"issn":["2751-4838"]},"volume":2,"author":[{"last_name":"Banerjee","first_name":"Tamajit","full_name":"Banerjee, Tamajit"},{"last_name":"Majumdar","first_name":"Rupak","full_name":"Majumdar, Rupak"},{"orcid":"0000-0001-9864-7475","last_name":"Mallik","first_name":"Kaushik","id":"0834ff3c-6d72-11ec-94e0-b5b0a4fb8598","full_name":"Mallik, Kaushik"},{"full_name":"Schmuck, Anne-Kathrin","last_name":"Schmuck","first_name":"Anne-Kathrin"},{"full_name":"Soudjani, Sadegh","first_name":"Sadegh","last_name":"Soudjani"}],"external_id":{"arxiv":["2202.07480"]}},{"oa":1,"arxiv":1,"title":"Deep neural collapse is provably optimal for the deep unconstrained features model","quality_controlled":"1","main_file_link":[{"url":" https://doi.org/10.48550/arXiv.2305.13165","open_access":"1"}],"department":[{"_id":"MaMo"},{"_id":"ChLa"}],"acknowledgement":"M. M. is partially supported by the 2019 Lopez-Loreta Prize. The authors would like to thank Eugenia Iofinova, Bernd Prach and Simone Bombari for valuable feedback on the manuscript.","day":"15","year":"2023","language":[{"iso":"eng"}],"date_created":"2024-02-02T11:17:41Z","status":"public","publication":"37th Annual Conference on Neural Information Processing Systems","article_processing_charge":"No","publication_status":"inpress","date_published":"2023-12-15T00:00:00Z","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"12","_id":"14921","type":"conference","citation":{"ama":"Súkeník P, Mondelli M, Lampert C. Deep neural collapse is provably optimal for the deep unconstrained features model. In: <i>37th Annual Conference on Neural Information Processing Systems</i>.","ieee":"P. Súkeník, M. Mondelli, and C. Lampert, “Deep neural collapse is provably optimal for the deep unconstrained features model,” in <i>37th Annual Conference on Neural Information Processing Systems</i>, New Orleans, LA, United States.","mla":"Súkeník, Peter, et al. “Deep Neural Collapse Is Provably Optimal for the Deep Unconstrained Features Model.” <i>37th Annual Conference on Neural Information Processing Systems</i>.","apa":"Súkeník, P., Mondelli, M., &#38; Lampert, C. (n.d.). Deep neural collapse is provably optimal for the deep unconstrained features model. In <i>37th Annual Conference on Neural Information Processing Systems</i>. New Orleans, LA, United States.","short":"P. Súkeník, M. Mondelli, C. Lampert, in:, 37th Annual Conference on Neural Information Processing Systems, n.d.","ista":"Súkeník P, Mondelli M, Lampert C. Deep neural collapse is provably optimal for the deep unconstrained features model. 37th Annual Conference on Neural Information Processing Systems. NeurIPS: Neural Information Processing Systems, NeurIPS, .","chicago":"Súkeník, Peter, Marco Mondelli, and Christoph Lampert. “Deep Neural Collapse Is Provably Optimal for the Deep Unconstrained Features Model.” In <i>37th Annual Conference on Neural Information Processing Systems</i>, n.d."},"date_updated":"2024-09-10T13:03:19Z","abstract":[{"lang":"eng","text":"Neural collapse (NC) refers to the surprising structure of the last layer of deep neural networks in the terminal phase of gradient descent training. Recently, an increasing amount of experimental evidence has pointed to the propagation of NC to earlier layers of neural networks. However, while the NC in the last layer is well studied theoretically, much less is known about its multi-layered counterpart - deep neural collapse (DNC). In particular, existing work focuses either on linear layers or only on the last two layers at the price of an extra assumption. Our paper fills this gap by generalizing the established analytical framework for NC - the unconstrained features model - to multiple non-linear layers. Our key technical contribution is to show that, in a deep unconstrained features model, the unique global optimum for binary classification exhibits all the properties typical of DNC. This explains the existing experimental evidence of DNC. We also empirically show that (i) by optimizing deep unconstrained features models via gradient descent, the resulting solution agrees well with our theory, and (ii) trained networks recover the unconstrained features suitable for the occurrence of DNC, thus supporting the validity of this modeling principle."}],"oa_version":"Preprint","project":[{"_id":"059876FA-7A3F-11EA-A408-12923DDC885E","name":"Prix Lopez-Loretta 2019 - Marco Mondelli"}],"conference":{"start_date":"2023-12-10","name":"NeurIPS: Neural Information Processing Systems","end_date":"2023-12-16","location":"New Orleans, LA, United States"},"author":[{"full_name":"Súkeník, Peter","first_name":"Peter","last_name":"Súkeník","id":"d64d6a8d-eb8e-11eb-b029-96fd216dec3c"},{"full_name":"Mondelli, Marco","id":"27EB676C-8706-11E9-9510-7717E6697425","last_name":"Mondelli","orcid":"0000-0002-3242-7020","first_name":"Marco"},{"full_name":"Lampert, Christoph","first_name":"Christoph","orcid":"0000-0001-8622-7887","last_name":"Lampert","id":"40C20FD2-F248-11E8-B48F-1D18A9856A87"}],"alternative_title":["NeurIPS"],"external_id":{"arxiv":["2305.13165"]}},{"type":"conference","date_updated":"2024-02-14T14:24:25Z","abstract":[{"text":"We propose a novel approach to concentration for non-independent random variables. The main idea is to ``pretend'' that the random variables are independent and pay a multiplicative price measuring how far they are from actually being independent. This price is encapsulated in the Hellinger integral between the joint and the product of the marginals, which is then upper bounded leveraging tensorisation properties. Our bounds represent a natural generalisation of concentration inequalities in the presence of dependence: we recover exactly the classical bounds (McDiarmid's inequality) when the random variables are independent. Furthermore, in a ``large deviations'' regime, we obtain the same decay in the probability as for the independent case, even when the random variables display non-trivial dependencies. To show this, we consider a number of applications of interest. First, we provide a bound for Markov chains with finite state space. Then, we consider the Simple Symmetric Random Walk, which is a non-contracting Markov chain, and a non-Markovian setting in which the stochastic process depends on its entire past. To conclude, we propose an application to Markov Chain Monte Carlo methods, where our approach leads to an improved lower bound on the minimum burn-in period required to reach a certain accuracy. In all of these settings, we provide a regime of parameters in which our bound fares better than what the state of the art can provide.","lang":"eng"}],"citation":{"chicago":"Esposito, Amedeo Roberto, and Marco Mondelli. “Concentration without Independence via Information Measures.” In <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>. IEEE, n.d. <a href=\"https://doi.org/10.1109/isit54713.2023.10206899\">https://doi.org/10.1109/isit54713.2023.10206899</a>.","ista":"Esposito AR, Mondelli M. Concentration without independence via information measures. Proceedings of 2023 IEEE International Symposium on Information Theory. ISIT: IEEE International Symposium on Information Theory.","short":"A.R. Esposito, M. Mondelli, in:, Proceedings of 2023 IEEE International Symposium on Information Theory, IEEE, n.d.","apa":"Esposito, A. R., &#38; Mondelli, M. (n.d.). Concentration without independence via information measures. In <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>. Taipei, Taiwan: IEEE. <a href=\"https://doi.org/10.1109/isit54713.2023.10206899\">https://doi.org/10.1109/isit54713.2023.10206899</a>","mla":"Esposito, Amedeo Roberto, and Marco Mondelli. “Concentration without Independence via Information Measures.” <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>, IEEE, doi:<a href=\"https://doi.org/10.1109/isit54713.2023.10206899\">10.1109/isit54713.2023.10206899</a>.","ieee":"A. R. Esposito and M. Mondelli, “Concentration without independence via information measures,” in <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>, Taipei, Taiwan.","ama":"Esposito AR, Mondelli M. Concentration without independence via information measures. In: <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>. IEEE. doi:<a href=\"https://doi.org/10.1109/isit54713.2023.10206899\">10.1109/isit54713.2023.10206899</a>"},"_id":"14922","publication_status":"inpress","date_published":"2023-06-30T00:00:00Z","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"06","conference":{"start_date":"2023-06-25","name":"ISIT: IEEE International Symposium on Information Theory","end_date":"2023-06-30","location":"Taipei, Taiwan"},"doi":"10.1109/isit54713.2023.10206899","author":[{"full_name":"Esposito, Amedeo Roberto","last_name":"Esposito","first_name":"Amedeo Roberto","id":"9583e921-e1ad-11ec-9862-cef099626dc9"},{"full_name":"Mondelli, Marco","id":"27EB676C-8706-11E9-9510-7717E6697425","orcid":"0000-0002-3242-7020","last_name":"Mondelli","first_name":"Marco"}],"external_id":{"arxiv":["2303.07245"]},"project":[{"_id":"059876FA-7A3F-11EA-A408-12923DDC885E","name":"Prix Lopez-Loretta 2019 - Marco Mondelli"}],"oa_version":"Preprint","day":"30","department":[{"_id":"MaMo"}],"acknowledgement":"The authors are partially supported by the 2019 Lopez-Loreta Prize. They would also like to thank Professor Jan Maas for providing valuable suggestions and comments on an early version of the work.","title":"Concentration without independence via information measures","quality_controlled":"1","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2303.07245","open_access":"1"}],"oa":1,"arxiv":1,"article_processing_charge":"No","publication":"Proceedings of 2023 IEEE International Symposium on Information Theory","publisher":"IEEE","date_created":"2024-02-02T11:18:40Z","status":"public","year":"2023","language":[{"iso":"eng"}]},{"publication":"Proceedings of 2023 IEEE International Symposium on Information Theory","article_processing_charge":"No","publisher":"IEEE","date_created":"2024-02-02T11:20:39Z","status":"public","year":"2023","language":[{"iso":"eng"}],"day":"30","department":[{"_id":"MaMo"}],"title":"Mismatched estimation of non-symmetric rank-one matrices corrupted by structured noise","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2302.03306","open_access":"1"}],"quality_controlled":"1","oa":1,"arxiv":1,"doi":"10.1109/isit54713.2023.10206671","conference":{"end_date":"2023-06-30","start_date":"2023-06-25","name":"ISIT: IEEE International Symposium on Information Theory","location":"Taipei, Taiwan"},"author":[{"full_name":"Fu, Teng","last_name":"Fu","first_name":"Teng"},{"last_name":"Liu","first_name":"YuHao","full_name":"Liu, YuHao"},{"full_name":"Barbier, Jean","last_name":"Barbier","first_name":"Jean"},{"full_name":"Mondelli, Marco","id":"27EB676C-8706-11E9-9510-7717E6697425","first_name":"Marco","orcid":"0000-0002-3242-7020","last_name":"Mondelli"},{"full_name":"Liang, ShanSuo","first_name":"ShanSuo","last_name":"Liang"},{"full_name":"Hou, TianQi","last_name":"Hou","first_name":"TianQi"}],"external_id":{"arxiv":["2302.03306"]},"oa_version":"Preprint","type":"conference","citation":{"ama":"Fu T, Liu Y, Barbier J, Mondelli M, Liang S, Hou T. Mismatched estimation of non-symmetric rank-one matrices corrupted by structured noise. In: <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>. IEEE. doi:<a href=\"https://doi.org/10.1109/isit54713.2023.10206671\">10.1109/isit54713.2023.10206671</a>","ieee":"T. Fu, Y. Liu, J. Barbier, M. Mondelli, S. Liang, and T. Hou, “Mismatched estimation of non-symmetric rank-one matrices corrupted by structured noise,” in <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>, Taipei, Taiwan.","mla":"Fu, Teng, et al. “Mismatched Estimation of Non-Symmetric Rank-One Matrices Corrupted by Structured Noise.” <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>, IEEE, doi:<a href=\"https://doi.org/10.1109/isit54713.2023.10206671\">10.1109/isit54713.2023.10206671</a>.","apa":"Fu, T., Liu, Y., Barbier, J., Mondelli, M., Liang, S., &#38; Hou, T. (n.d.). Mismatched estimation of non-symmetric rank-one matrices corrupted by structured noise. In <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>. Taipei, Taiwan: IEEE. <a href=\"https://doi.org/10.1109/isit54713.2023.10206671\">https://doi.org/10.1109/isit54713.2023.10206671</a>","short":"T. Fu, Y. Liu, J. Barbier, M. Mondelli, S. Liang, T. Hou, in:, Proceedings of 2023 IEEE International Symposium on Information Theory, IEEE, n.d.","ista":"Fu T, Liu Y, Barbier J, Mondelli M, Liang S, Hou T. Mismatched estimation of non-symmetric rank-one matrices corrupted by structured noise. Proceedings of 2023 IEEE International Symposium on Information Theory. ISIT: IEEE International Symposium on Information Theory.","chicago":"Fu, Teng, YuHao Liu, Jean Barbier, Marco Mondelli, ShanSuo Liang, and TianQi Hou. “Mismatched Estimation of Non-Symmetric Rank-One Matrices Corrupted by Structured Noise.” In <i>Proceedings of 2023 IEEE International Symposium on Information Theory</i>. IEEE, n.d. <a href=\"https://doi.org/10.1109/isit54713.2023.10206671\">https://doi.org/10.1109/isit54713.2023.10206671</a>."},"abstract":[{"lang":"eng","text":"We study the performance of a Bayesian statistician who estimates a rank-one signal corrupted by non-symmetric rotationally invariant noise with a generic distribution of singular values. As the signal-to-noise ratio and the noise structure are unknown, a Gaussian setup is incorrectly assumed. We derive the exact analytic expression for the error of the mismatched Bayes estimator and also provide the analysis of an approximate message passing (AMP) algorithm. The first result exploits the asymptotic behavior of spherical integrals for rectangular matrices and of low-rank matrix perturbations; the second one relies on the design and analysis of an auxiliary AMP. The numerical experiments show that there is a performance gap between the AMP and Bayes estimators, which is due to the incorrect estimation of the signal norm."}],"date_updated":"2024-02-14T14:34:03Z","_id":"14923","publication_status":"inpress","date_published":"2023-06-30T00:00:00Z","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"06"},{"citation":{"apa":"Wu, D., Kungurtsev, V., &#38; Mondelli, M. (2023). Mean-field analysis for heavy ball methods: Dropout-stability, connectivity, and global convergence. In <i>Transactions on Machine Learning Research</i>. ML Research Press.","short":"D. Wu, V. Kungurtsev, M. Mondelli, in:, Transactions on Machine Learning Research, ML Research Press, 2023.","ista":"Wu D, Kungurtsev V, Mondelli M. 2023. Mean-field analysis for heavy ball methods: Dropout-stability, connectivity, and global convergence. Transactions on Machine Learning Research. , TMLR, .","chicago":"Wu, Diyuan, Vyacheslav Kungurtsev, and Marco Mondelli. “Mean-Field Analysis for Heavy Ball Methods: Dropout-Stability, Connectivity, and Global Convergence.” In <i>Transactions on Machine Learning Research</i>. ML Research Press, 2023.","ama":"Wu D, Kungurtsev V, Mondelli M. Mean-field analysis for heavy ball methods: Dropout-stability, connectivity, and global convergence. In: <i>Transactions on Machine Learning Research</i>. ML Research Press; 2023.","ieee":"D. Wu, V. Kungurtsev, and M. Mondelli, “Mean-field analysis for heavy ball methods: Dropout-stability, connectivity, and global convergence,” in <i>Transactions on Machine Learning Research</i>, 2023.","mla":"Wu, Diyuan, et al. “Mean-Field Analysis for Heavy Ball Methods: Dropout-Stability, Connectivity, and Global Convergence.” <i>Transactions on Machine Learning Research</i>, ML Research Press, 2023."},"date_updated":"2024-09-10T13:03:20Z","has_accepted_license":"1","type":"conference","month":"02","date_published":"2023-02-28T00:00:00Z","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)","image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)"},"alternative_title":["TMLR"],"oa_version":"Published Version","acknowledgement":"D. Wu and M. Mondelli are partially supported by the 2019 Lopez-Loreta Prize. V. Kungurtsev was supported by the OP VVV project CZ.02.1.01/0.0/0.0/16_019/0000765 \"Research Center for Informatics\".","title":"Mean-field analysis for heavy ball methods: Dropout-stability, connectivity, and global convergence","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2210.06819","open_access":"1"}],"publication":"Transactions on Machine Learning Research","article_processing_charge":"No","language":[{"iso":"eng"}],"year":"2023","abstract":[{"lang":"eng","text":"The stochastic heavy ball method (SHB), also known as stochastic gradient descent (SGD) with Polyak's momentum, is widely used in training neural networks. However, despite the remarkable success of such algorithm in practice, its theoretical characterization remains limited. In this paper, we focus on neural networks with two and three layers and provide a rigorous understanding of the properties of the solutions found by SHB: \\emph{(i)} stability after dropping out part of the neurons, \\emph{(ii)} connectivity along a low-loss path, and \\emph{(iii)} convergence to the global optimum.\r\nTo achieve this goal, we take a mean-field view and relate the SHB dynamics to a certain partial differential equation in the limit of large network widths. This mean-field perspective has inspired a recent line of work focusing on SGD while, in contrast, our paper considers an algorithm with momentum. More specifically, after proving existence and uniqueness of the limit differential equations, we show convergence to the global optimum and give a quantitative bound between the mean-field limit and the SHB dynamics of a finite-width network. Armed with this last bound, we are able to establish the dropout-stability and connectivity of SHB solutions."}],"_id":"14924","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","publication_status":"published","author":[{"id":"1a5914c2-896a-11ed-bdf8-fb80621a0635","last_name":"Wu","first_name":"Diyuan","full_name":"Wu, Diyuan"},{"full_name":"Kungurtsev, Vyacheslav","first_name":"Vyacheslav","last_name":"Kungurtsev"},{"id":"27EB676C-8706-11E9-9510-7717E6697425","last_name":"Mondelli","orcid":"0000-0002-3242-7020","first_name":"Marco","full_name":"Mondelli, Marco"}],"external_id":{"arxiv":["2210.06819"]},"project":[{"_id":"059876FA-7A3F-11EA-A408-12923DDC885E","name":"Prix Lopez-Loretta 2019 - Marco Mondelli"}],"day":"28","department":[{"_id":"MaMo"}],"quality_controlled":"1","arxiv":1,"oa":1,"publisher":"ML Research Press","status":"public","date_created":"2024-02-02T11:21:56Z"},{"status":"public","date_created":"2024-02-07T14:28:34Z","language":[{"iso":"eng"}],"year":"2023","publication":"arXiv","article_processing_charge":"No","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2311.04056","open_access":"1"}],"title":"Multi-view causal representation learning with partial observability","arxiv":1,"oa":1,"day":"07","acknowledgement":"This work was initiated at the Second Bellairs Workshop on Causality held at the Bellairs Research Institute, January 6–13, 2022; we thank all workshop participants for providing a stimulating research environment. Further, we thank Cian Eastwood, Luigi Gresele, Stefano Soatto, Marco Bagatella, and A. René Geist for helpful discussion. GM is a member of the Machine Learning Cluster of Excellence, EXC number 2064/1 – Project number 390727645. JvK and GM acknowledge support from the German Federal Ministry of Education and Research (BMBF) through the Tübingen AI Center (FKZ: 01IS18039B). The research of DX and SM was supported by the Air Force Office of Scientific Research under award number FA8655-22-1-7155. Any opinions, findings, and conclusions or recommendations expressed in\r\nthis material are those of the author(s) and do not necessarily reflect the views of the United States Air Force. We also thank SURF for the support in using the Dutch National Supercomputer Snellius. DY was supported by an Amazon fellowship and the International Max Planck Research School for Intelligent Systems (IMPRS-IS). Work done outside of Amazon. SL was supported by an IVADO excellence PhD scholarship and by Samsung Electronics Co., Ldt.","department":[{"_id":"FrLo"}],"oa_version":"Preprint","author":[{"first_name":"Dingling","last_name":"Yao","id":"d3e02e50-48a8-11ee-8f62-c108061797fa","full_name":"Yao, Dingling"},{"full_name":"Xu, Danru","first_name":"Danru","last_name":"Xu"},{"full_name":"Lachapelle, Sébastien","last_name":"Lachapelle","first_name":"Sébastien"},{"first_name":"Sara","last_name":"Magliacane","full_name":"Magliacane, Sara"},{"full_name":"Taslakian, Perouz","last_name":"Taslakian","first_name":"Perouz"},{"full_name":"Martius, Georg","first_name":"Georg","last_name":"Martius"},{"first_name":"Julius von","last_name":"Kügelgen","full_name":"Kügelgen, Julius von"},{"full_name":"Locatello, Francesco","id":"26cfd52f-2483-11ee-8040-88983bcc06d4","first_name":"Francesco","last_name":"Locatello","orcid":"0000-0002-4850-0683"}],"external_id":{"arxiv":["2311.04056"]},"doi":"10.48550/arXiv.2311.04056","month":"11","date_published":"2023-11-07T00:00:00Z","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","publication_status":"submitted","date_updated":"2024-02-12T08:07:33Z","abstract":[{"text":"We present a unified framework for studying the identifiability of\r\nrepresentations learned from simultaneously observed views, such as different\r\ndata modalities. We allow a partially observed setting in which each view\r\nconstitutes a nonlinear mixture of a subset of underlying latent variables,\r\nwhich can be causally related. We prove that the information shared across all\r\nsubsets of any number of views can be learned up to a smooth bijection using\r\ncontrastive learning and a single encoder per view. We also provide graphical\r\ncriteria indicating which latent variables can be identified through a simple\r\nset of rules, which we refer to as identifiability algebra. Our general\r\nframework and theoretical results unify and extend several previous works on\r\nmulti-view nonlinear ICA, disentanglement, and causal representation learning.\r\nWe experimentally validate our claims on numerical, image, and multi-modal data\r\nsets. Further, we demonstrate that the performance of prior methods is\r\nrecovered in different special cases of our setup. Overall, we find that access\r\nto multiple partial views enables us to identify a more fine-grained\r\nrepresentation, under the generally milder assumption of partial observability.","lang":"eng"}],"citation":{"chicago":"Yao, Dingling, Danru Xu, Sébastien Lachapelle, Sara Magliacane, Perouz Taslakian, Georg Martius, Julius von Kügelgen, and Francesco Locatello. “Multi-View Causal Representation Learning with Partial Observability.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2311.04056\">https://doi.org/10.48550/arXiv.2311.04056</a>.","ista":"Yao D, Xu D, Lachapelle S, Magliacane S, Taslakian P, Martius G, Kügelgen J von, Locatello F. Multi-view causal representation learning with partial observability. arXiv, 2311.04056.","short":"D. Yao, D. Xu, S. Lachapelle, S. Magliacane, P. Taslakian, G. Martius, J. von Kügelgen, F. Locatello, ArXiv (n.d.).","apa":"Yao, D., Xu, D., Lachapelle, S., Magliacane, S., Taslakian, P., Martius, G., … Locatello, F. (n.d.). Multi-view causal representation learning with partial observability. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2311.04056\">https://doi.org/10.48550/arXiv.2311.04056</a>","mla":"Yao, Dingling, et al. “Multi-View Causal Representation Learning with Partial Observability.” <i>ArXiv</i>, 2311.04056, doi:<a href=\"https://doi.org/10.48550/arXiv.2311.04056\">10.48550/arXiv.2311.04056</a>.","ieee":"D. Yao <i>et al.</i>, “Multi-view causal representation learning with partial observability,” <i>arXiv</i>. .","ama":"Yao D, Xu D, Lachapelle S, et al. Multi-view causal representation learning with partial observability. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2311.04056\">10.48550/arXiv.2311.04056</a>"},"type":"preprint","article_number":"2311.04056","_id":"14946"},{"oa_version":"Preprint","author":[{"first_name":"Avinash","last_name":"Kori","full_name":"Kori, Avinash"},{"full_name":"Locatello, Francesco","id":"26cfd52f-2483-11ee-8040-88983bcc06d4","orcid":"0000-0002-4850-0683","last_name":"Locatello","first_name":"Francesco"},{"full_name":"Ribeiro, Fabio De Sousa","last_name":"Ribeiro","first_name":"Fabio De Sousa"},{"full_name":"Toni, Francesca","last_name":"Toni","first_name":"Francesca"},{"full_name":"Glocker, Ben","last_name":"Glocker","first_name":"Ben"}],"external_id":{"arxiv":["2307.09437"]},"doi":"10.48550/arXiv.2307.09437","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","date_published":"2023-07-18T00:00:00Z","month":"07","publication_status":"submitted","article_number":"2307.09437","_id":"14948","citation":{"chicago":"Kori, Avinash, Francesco Locatello, Fabio De Sousa Ribeiro, Francesca Toni, and Ben Glocker. “Grounded Object Centric Learning.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2307.09437\">https://doi.org/10.48550/arXiv.2307.09437</a>.","ista":"Kori A, Locatello F, Ribeiro FDS, Toni F, Glocker B. Grounded object centric learning. arXiv, 2307.09437.","short":"A. Kori, F. Locatello, F.D.S. Ribeiro, F. Toni, B. Glocker, ArXiv (n.d.).","apa":"Kori, A., Locatello, F., Ribeiro, F. D. S., Toni, F., &#38; Glocker, B. (n.d.). Grounded object centric learning. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2307.09437\">https://doi.org/10.48550/arXiv.2307.09437</a>","mla":"Kori, Avinash, et al. “Grounded Object Centric Learning.” <i>ArXiv</i>, 2307.09437, doi:<a href=\"https://doi.org/10.48550/arXiv.2307.09437\">10.48550/arXiv.2307.09437</a>.","ieee":"A. Kori, F. Locatello, F. D. S. Ribeiro, F. Toni, and B. Glocker, “Grounded object centric learning,” <i>arXiv</i>. .","ama":"Kori A, Locatello F, Ribeiro FDS, Toni F, Glocker B. Grounded object centric learning. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2307.09437\">10.48550/arXiv.2307.09437</a>"},"abstract":[{"lang":"eng","text":"The extraction of modular object-centric representations for downstream tasks\r\nis an emerging area of research. Learning grounded representations of objects\r\nthat are guaranteed to be stable and invariant promises robust performance\r\nacross different tasks and environments. Slot Attention (SA) learns\r\nobject-centric representations by assigning objects to \\textit{slots}, but\r\npresupposes a \\textit{single} distribution from which all slots are randomly\r\ninitialised. This results in an inability to learn \\textit{specialized} slots\r\nwhich bind to specific object types and remain invariant to identity-preserving\r\nchanges in object appearance. To address this, we present\r\n\\emph{\\textsc{Co}nditional \\textsc{S}lot \\textsc{A}ttention} (\\textsc{CoSA})\r\nusing a novel concept of \\emph{Grounded Slot Dictionary} (GSD) inspired by\r\nvector quantization. Our proposed GSD comprises (i) canonical object-level\r\nproperty vectors and (ii) parametric Gaussian distributions, which define a\r\nprior over the slots. We demonstrate the benefits of our method in multiple\r\ndownstream tasks such as scene generation, composition, and task adaptation,\r\nwhilst remaining competitive with SA in popular object discovery benchmarks."}],"date_updated":"2024-02-12T08:13:12Z","type":"preprint","language":[{"iso":"eng"}],"year":"2023","status":"public","date_created":"2024-02-07T14:47:04Z","article_processing_charge":"No","publication":"arXiv","arxiv":1,"oa":1,"title":"Grounded object centric learning","main_file_link":[{"open_access":"1","url":"https://doi.org/10.48550/arXiv.2307.09437"}],"acknowledgement":"This work was supported by supported by UKRI (grant agreement no. EP/S023356/1), in the UKRI\r\nCentre for Doctoral Training in Safe and Trusted AI via A. Kori.","department":[{"_id":"FrLo"}],"day":"18"},{"abstract":[{"text":"Many approaches have been proposed to use diffusion models to augment training datasets for downstream tasks, such as classification. However, diffusion models are themselves trained on large datasets, often with noisy annotations, and it remains an open question to which extent these models contribute to downstream classification performance. In particular, it remains unclear if they generalize enough to improve over directly using the additional data of their pre-training process for augmentation. We systematically evaluate a range of existing methods to generate images from diffusion models and study new extensions to assess their benefit for data augmentation. Personalizing diffusion models towards the target data outperforms simpler prompting strategies. However, using the pre-training data of the diffusion model alone, via a simple nearest-neighbor retrieval procedure, leads to even stronger downstream performance. Our study explores the potential of diffusion models in generating new training data, and surprisingly finds that these sophisticated models are not yet able to beat a simple and strong image retrieval baseline on simple downstream vision tasks.","lang":"eng"}],"_id":"14949","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","publication_status":"published","author":[{"last_name":"Burg","first_name":"Max","full_name":"Burg, Max"},{"full_name":"Wenzel, Florian","last_name":"Wenzel","first_name":"Florian"},{"full_name":"Zietlow, Dominik","last_name":"Zietlow","first_name":"Dominik"},{"first_name":"Max","last_name":"Horn","full_name":"Horn, Max"},{"full_name":"Makansi, Osama","first_name":"Osama","last_name":"Makansi"},{"first_name":"Francesco","orcid":"0000-0002-4850-0683","last_name":"Locatello","id":"26cfd52f-2483-11ee-8040-88983bcc06d4","full_name":"Locatello, Francesco"},{"first_name":"Chris","last_name":"Russell","full_name":"Russell, Chris"}],"publication_identifier":{"eissn":["2835-8856"]},"day":"10","department":[{"_id":"FrLo"}],"quality_controlled":"1","oa":1,"file":[{"file_id":"14950","relation":"main_file","content_type":"application/pdf","file_size":27325153,"date_updated":"2024-02-07T14:57:32Z","creator":"ptazenko","date_created":"2024-02-07T14:57:32Z","checksum":"af87ddea7908923426365347b9c87ba7","access_level":"open_access","file_name":"Burg_et_al_2023_Image_retrieval_outperforms.pdf"}],"publisher":"ML Research Press","status":"public","date_created":"2024-02-07T14:57:39Z","citation":{"short":"M. Burg, F. Wenzel, D. Zietlow, M. Horn, O. Makansi, F. Locatello, C. Russell, Journal of Machine Learning Research (2023).","apa":"Burg, M., Wenzel, F., Zietlow, D., Horn, M., Makansi, O., Locatello, F., &#38; Russell, C. (2023). Image retrieval outperforms diffusion models on data augmentation. <i>Journal of Machine Learning Research</i>. ML Research Press.","chicago":"Burg, Max, Florian Wenzel, Dominik Zietlow, Max Horn, Osama Makansi, Francesco Locatello, and Chris Russell. “Image Retrieval Outperforms Diffusion Models on Data Augmentation.” <i>Journal of Machine Learning Research</i>. ML Research Press, 2023.","ista":"Burg M, Wenzel F, Zietlow D, Horn M, Makansi O, Locatello F, Russell C. 2023. Image retrieval outperforms diffusion models on data augmentation. Journal of Machine Learning Research.","ama":"Burg M, Wenzel F, Zietlow D, et al. Image retrieval outperforms diffusion models on data augmentation. <i>Journal of Machine Learning Research</i>. 2023.","mla":"Burg, Max, et al. “Image Retrieval Outperforms Diffusion Models on Data Augmentation.” <i>Journal of Machine Learning Research</i>, ML Research Press, 2023.","ieee":"M. Burg <i>et al.</i>, “Image retrieval outperforms diffusion models on data augmentation,” <i>Journal of Machine Learning Research</i>. ML Research Press, 2023."},"has_accepted_license":"1","date_updated":"2024-02-12T08:30:21Z","type":"journal_article","date_published":"2023-12-10T00:00:00Z","month":"12","ddc":["000"],"tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)","image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)"},"alternative_title":["TMLR"],"article_type":"original","oa_version":"Published Version","file_date_updated":"2024-02-07T14:57:32Z","acknowledgement":"The authors would like to thank Varad Gunjal and Vishaal Udandarao. MFB thanks the International Max Planck Research School for Intelligent Systems (IMPRS-IS).","main_file_link":[{"url":"https://openreview.net/forum?id=xflYdGZMpv","open_access":"1"}],"title":"Image retrieval outperforms diffusion models on data augmentation","article_processing_charge":"No","publication":"Journal of Machine Learning Research","language":[{"iso":"eng"}],"year":"2023"},{"year":"2023","language":[{"iso":"eng"}],"status":"public","date_created":"2024-02-07T15:08:55Z","publication":"arXiv","article_processing_charge":"No","arxiv":1,"oa":1,"main_file_link":[{"open_access":"1","url":"https://doi.org/10.48550/arXiv.2311.00664"}],"title":"Latent space translation via semantic alignment","acknowledgement":"This work is supported by the ERC grant no.802554 (SPECGEO), PRIN 2020 project no.2020TA3K9N (LEGO.AI), and PNRR MUR project PE0000013-FAIR. Francesco\r\nLocatello did not contribute to this work at Amazon.","department":[{"_id":"FrLo"}],"day":"01","oa_version":"Preprint","external_id":{"arxiv":["2311.00664"]},"author":[{"first_name":"Valentino","last_name":"Maiorca","full_name":"Maiorca, Valentino"},{"last_name":"Moschella","first_name":"Luca","full_name":"Moschella, Luca"},{"full_name":"Norelli, Antonio","first_name":"Antonio","last_name":"Norelli"},{"last_name":"Fumero","first_name":"Marco","full_name":"Fumero, Marco"},{"first_name":"Francesco","last_name":"Locatello","orcid":"0000-0002-4850-0683","id":"26cfd52f-2483-11ee-8040-88983bcc06d4","full_name":"Locatello, Francesco"},{"full_name":"Rodolà, Emanuele","first_name":"Emanuele","last_name":"Rodolà"}],"doi":"10.48550/arXiv.2311.00664","month":"11","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","date_published":"2023-11-01T00:00:00Z","publication_status":"submitted","article_number":"2311.00664","_id":"14952","date_updated":"2024-02-12T09:40:23Z","citation":{"ama":"Maiorca V, Moschella L, Norelli A, Fumero M, Locatello F, Rodolà E. Latent space translation via semantic alignment. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2311.00664\">10.48550/arXiv.2311.00664</a>","ieee":"V. Maiorca, L. Moschella, A. Norelli, M. Fumero, F. Locatello, and E. Rodolà, “Latent space translation via semantic alignment,” <i>arXiv</i>. .","mla":"Maiorca, Valentino, et al. “Latent Space Translation via Semantic Alignment.” <i>ArXiv</i>, 2311.00664, doi:<a href=\"https://doi.org/10.48550/arXiv.2311.00664\">10.48550/arXiv.2311.00664</a>.","apa":"Maiorca, V., Moschella, L., Norelli, A., Fumero, M., Locatello, F., &#38; Rodolà, E. (n.d.). Latent space translation via semantic alignment. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2311.00664\">https://doi.org/10.48550/arXiv.2311.00664</a>","short":"V. Maiorca, L. Moschella, A. Norelli, M. Fumero, F. Locatello, E. Rodolà, ArXiv (n.d.).","ista":"Maiorca V, Moschella L, Norelli A, Fumero M, Locatello F, Rodolà E. Latent space translation via semantic alignment. arXiv, 2311.00664.","chicago":"Maiorca, Valentino, Luca Moschella, Antonio Norelli, Marco Fumero, Francesco Locatello, and Emanuele Rodolà. “Latent Space Translation via Semantic Alignment.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2311.00664\">https://doi.org/10.48550/arXiv.2311.00664</a>."},"abstract":[{"lang":"eng","text":"While different neural models often exhibit latent spaces that are alike when exposed to semantically related data, this intrinsic similarity is not always immediately discernible. Towards a better understanding of this phenomenon, our work shows how representations learned from these neural modules can be translated between different pre-trained networks via simpler transformations than previously thought. An advantage of this approach is the ability to\r\nestimate these transformations using standard, well-understood algebraic procedures that have closed-form solutions. Our method directly estimates a transformation between two given latent spaces, thereby enabling effective stitching of encoders and decoders without additional training. We extensively validate the adaptability of this translation procedure in different\r\nexperimental settings: across various trainings, domains, architectures (e.g., ResNet, CNN, ViT), and in multiple downstream tasks (classification, reconstruction). Notably, we show how it is possible to zero-shot stitch text encoders and vision decoders, or vice-versa, yielding surprisingly good classification performance in this multimodal setting."}],"type":"preprint"},{"doi":"10.48550/arXiv.2310.18123","external_id":{"arxiv":["2310.18123"]},"author":[{"first_name":"Zhenyu","last_name":"Zhu","full_name":"Zhu, Zhenyu"},{"full_name":"Locatello, Francesco","id":"26cfd52f-2483-11ee-8040-88983bcc06d4","first_name":"Francesco","last_name":"Locatello","orcid":"0000-0002-4850-0683"},{"full_name":"Cevher, Volkan","last_name":"Cevher","first_name":"Volkan"}],"oa_version":"Preprint","_id":"14953","article_number":"2310.18123","type":"preprint","abstract":[{"lang":"eng","text":"This paper provides statistical sample complexity bounds for score-matching and its applications in causal discovery. We demonstrate that accurate estimation of the score function is achievable by training a standard deep ReLU neural network using stochastic gradient descent. We establish bounds on the error rate of recovering causal relationships using the score-matching-based causal discovery method of Rolland et al. [2022], assuming a sufficiently good estimation of the score function. Finally, we analyze the upper bound of score-matching estimation within the score-based generative modeling, which has been applied for causal discovery but is also of independent interest within the domain of generative models."}],"date_updated":"2024-02-12T09:45:58Z","citation":{"ista":"Zhu Z, Locatello F, Cevher V. Sample complexity bounds for score-matching: Causal discovery and generative modeling. arXiv, 2310.18123.","chicago":"Zhu, Zhenyu, Francesco Locatello, and Volkan Cevher. “Sample Complexity Bounds for Score-Matching: Causal Discovery and Generative Modeling.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2310.18123\">https://doi.org/10.48550/arXiv.2310.18123</a>.","apa":"Zhu, Z., Locatello, F., &#38; Cevher, V. (n.d.). Sample complexity bounds for score-matching: Causal discovery and generative modeling. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2310.18123\">https://doi.org/10.48550/arXiv.2310.18123</a>","short":"Z. Zhu, F. Locatello, V. Cevher, ArXiv (n.d.).","ieee":"Z. Zhu, F. Locatello, and V. Cevher, “Sample complexity bounds for score-matching: Causal discovery and generative modeling,” <i>arXiv</i>. .","mla":"Zhu, Zhenyu, et al. “Sample Complexity Bounds for Score-Matching: Causal Discovery and Generative Modeling.” <i>ArXiv</i>, 2310.18123, doi:<a href=\"https://doi.org/10.48550/arXiv.2310.18123\">10.48550/arXiv.2310.18123</a>.","ama":"Zhu Z, Locatello F, Cevher V. Sample complexity bounds for score-matching: Causal discovery and generative modeling. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2310.18123\">10.48550/arXiv.2310.18123</a>"},"publication_status":"submitted","date_published":"2023-10-27T00:00:00Z","month":"10","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","article_processing_charge":"No","publication":"arXiv","year":"2023","language":[{"iso":"eng"}],"date_created":"2024-02-07T15:11:11Z","status":"public","department":[{"_id":"FrLo"}],"acknowledgement":"We are thankful to the reviewers for providing constructive feedback and Kun Zhang and Dominik Janzing for helpful discussion on the special case of deterministic children. This work was supported by Hasler Foundation Program: Hasler Responsible AI (project number 21043). This work was supported by the Swiss National Science Foundation (SNSF) under grant number 200021_205011. Francesco Locatello did not contribute to this work at Amazon. ","day":"27","oa":1,"arxiv":1,"title":"Sample complexity bounds for score-matching: Causal discovery and generative modeling","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2310.18123","open_access":"1"}]},{"_id":"14954","article_number":"2310.13387","type":"preprint","citation":{"ieee":"F. Montagna <i>et al.</i>, “Assumption violations in causal discovery and the robustness of score matching,” <i>arXiv</i>. .","mla":"Montagna, Francesco, et al. “Assumption Violations in Causal Discovery and the Robustness of Score Matching.” <i>ArXiv</i>, 2310.13387, doi:<a href=\"https://doi.org/10.48550/arXiv.2310.13387\">10.48550/arXiv.2310.13387</a>.","ama":"Montagna F, Mastakouri AA, Eulig E, et al. Assumption violations in causal discovery and the robustness of score matching. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2310.13387\">10.48550/arXiv.2310.13387</a>","ista":"Montagna F, Mastakouri AA, Eulig E, Noceti N, Rosasco L, Janzing D, Aragam B, Locatello F. Assumption violations in causal discovery and the robustness of score matching. arXiv, 2310.13387.","chicago":"Montagna, Francesco, Atalanti A. Mastakouri, Elias Eulig, Nicoletta Noceti, Lorenzo Rosasco, Dominik Janzing, Bryon Aragam, and Francesco Locatello. “Assumption Violations in Causal Discovery and the Robustness of Score Matching.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2310.13387\">https://doi.org/10.48550/arXiv.2310.13387</a>.","apa":"Montagna, F., Mastakouri, A. A., Eulig, E., Noceti, N., Rosasco, L., Janzing, D., … Locatello, F. (n.d.). Assumption violations in causal discovery and the robustness of score matching. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2310.13387\">https://doi.org/10.48550/arXiv.2310.13387</a>","short":"F. Montagna, A.A. Mastakouri, E. Eulig, N. Noceti, L. Rosasco, D. Janzing, B. Aragam, F. Locatello, ArXiv (n.d.)."},"date_updated":"2024-02-12T09:51:15Z","abstract":[{"lang":"eng","text":"When domain knowledge is limited and experimentation is restricted by ethical, financial, or time constraints, practitioners turn to observational causal discovery methods to recover the causal structure, exploiting the statistical properties of their data. Because causal discovery without further assumptions is an ill-posed problem, each algorithm comes with its own set of\r\nusually untestable assumptions, some of which are hard to meet in real datasets. Motivated by these considerations, this paper extensively benchmarks the empirical performance of recent causal discovery methods on observational i.i.d. data generated under different background conditions, allowing for violations of the critical assumptions required by each selected approach. Our experimental findings show that score matching-based methods demonstrate\r\nsurprising performance in the false positive and false negative rate of the inferred graph in these challenging scenarios, and we provide theoretical insights into their performance. This work is also the first effort to benchmark the stability of causal discovery algorithms with respect to the values of their hyperparameters. Finally, we hope this paper will set a new standard for the evaluation of causal discovery methods and can serve as an accessible entry point for practitioners interested in the field, highlighting the empirical implications of different algorithm choices."}],"publication_status":"submitted","month":"10","date_published":"2023-10-20T00:00:00Z","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","doi":"10.48550/arXiv.2310.13387","external_id":{"arxiv":["2310.13387"]},"author":[{"full_name":"Montagna, Francesco","last_name":"Montagna","first_name":"Francesco"},{"full_name":"Mastakouri, Atalanti A.","first_name":"Atalanti A.","last_name":"Mastakouri"},{"full_name":"Eulig, Elias","first_name":"Elias","last_name":"Eulig"},{"full_name":"Noceti, Nicoletta","first_name":"Nicoletta","last_name":"Noceti"},{"full_name":"Rosasco, Lorenzo","first_name":"Lorenzo","last_name":"Rosasco"},{"last_name":"Janzing","first_name":"Dominik","full_name":"Janzing, Dominik"},{"full_name":"Aragam, Bryon","last_name":"Aragam","first_name":"Bryon"},{"full_name":"Locatello, Francesco","last_name":"Locatello","orcid":"0000-0002-4850-0683","first_name":"Francesco","id":"26cfd52f-2483-11ee-8040-88983bcc06d4"}],"oa_version":"Preprint","department":[{"_id":"FrLo"}],"acknowledgement":"We thank Kun Zhang and Carl-Johann Simon-Gabriel for the insightful discussions. This work\r\nhas been supported by AFOSR, grant n. FA8655-20-1-7035. FM is supported by Programma\r\nOperativo Nazionale ricerca e innovazione 2014-2020. FM partially contributed to this work during an internship at Amazon Web Services with FL. FL partially contributed while at AWS.","day":"20","oa":1,"arxiv":1,"main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2310.13387","open_access":"1"}],"title":"Assumption violations in causal discovery and the robustness of score matching","publication":"arXiv","article_processing_charge":"No","year":"2023","language":[{"iso":"eng"}],"date_created":"2024-02-07T15:11:56Z","status":"public"},{"quality_controlled":"1","oa":1,"day":"05","department":[{"_id":"FrLo"}],"date_created":"2024-02-07T15:17:51Z","status":"public","file":[{"success":1,"checksum":"484efc27bda75ed6666044989695d9b6","date_created":"2024-02-13T08:50:53Z","file_name":"2023_CRL_Xu.pdf","access_level":"open_access","relation":"main_file","file_id":"14982","file_size":552357,"content_type":"application/pdf","creator":"dernst","date_updated":"2024-02-13T08:50:53Z"}],"publisher":"OpenReview","publication_status":"published","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","abstract":[{"lang":"eng","text":"Causal representation learning (CRL) aims at identifying high-level causal variables from low-level data, e.g. images. Current methods usually assume that all causal variables are captured in the high-dimensional observations. In this work, we focus on learning causal representations from data under partial observability, i.e., when some of the causal variables are not observed in the measurements, and the set of masked variables changes across the different samples. We introduce some initial theoretical results for identifying causal variables under partial observability by exploiting a sparsity regularizer, focusing in particular on the linear and piecewise linear mixing function case. We provide a theorem that allows us to identify the causal variables up to permutation and element-wise linear transformations in the linear case and a lemma that allows us to identify causal variables up to linear transformation in the piecewise case. Finally, we provide a conjecture that would allow us to identify the causal variables up to permutation and element-wise linear transformations also in the piecewise linear case. We test the theorem and conjecture on simulated data, showing the effectiveness of our method."}],"_id":"14958","author":[{"full_name":"Xu, Danru","first_name":"Danru","last_name":"Xu"},{"full_name":"Yao, Dingling","last_name":"Yao","first_name":"Dingling","id":"d3e02e50-48a8-11ee-8f62-c108061797fa"},{"full_name":"Lachapelle, Sebastien","first_name":"Sebastien","last_name":"Lachapelle"},{"full_name":"Taslakian, Perouz","last_name":"Taslakian","first_name":"Perouz"},{"full_name":"von Kügelgen, Julius","first_name":"Julius","last_name":"von Kügelgen"},{"full_name":"Locatello, Francesco","first_name":"Francesco","orcid":"0000-0002-4850-0683","last_name":"Locatello","id":"26cfd52f-2483-11ee-8040-88983bcc06d4"},{"first_name":"Sara","last_name":"Magliacane","full_name":"Magliacane, Sara"}],"title":"A sparsity principle for partially observable causal representation learning","main_file_link":[{"open_access":"1","url":"https://openreview.net/forum?id=Whr6uobelR"}],"acknowledgement":"This work was initiated at the Second Bellairs Workshop on Causality held at the Bellairs Research Institute, January 6–13, 2022; we thank all workshop participants for providing a stimulating research environment. The research of DX and SM was supported by the Air Force Office of Scientific Research under award number FA8655-22-1-7155. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the United States Air Force. We also thank SURF for the support in using the Dutch National Supercomputer Snellius. DY was supported by an Amazon fellowship and the International Max Planck Research School for Intelligent Systems (IMPRS-IS). Work done outside of Amazon. SL was supported by an IVADO excellence PhD scholarship and by Samsung Electronics Co., Ldt. JvK acknowledges support from the German Federal Ministry of Education and Research (BMBF)\r\nthrough the Tübingen AI Center (FKZ: 01IS18039B).\r\n","year":"2023","language":[{"iso":"eng"}],"article_processing_charge":"No","publication":"Causal Representation Learning Workshop at NeurIPS 2023","ddc":["000"],"month":"12","date_published":"2023-12-05T00:00:00Z","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)","image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)"},"type":"conference","has_accepted_license":"1","date_updated":"2024-02-13T08:59:27Z","citation":{"ama":"Xu D, Yao D, Lachapelle S, et al. A sparsity principle for partially observable causal representation learning. In: <i>Causal Representation Learning Workshop at NeurIPS 2023</i>. OpenReview; 2023.","mla":"Xu, Danru, et al. “A Sparsity Principle for Partially Observable Causal Representation Learning.” <i>Causal Representation Learning Workshop at NeurIPS 2023</i>, 54, OpenReview, 2023.","ieee":"D. Xu <i>et al.</i>, “A sparsity principle for partially observable causal representation learning,” in <i>Causal Representation Learning Workshop at NeurIPS 2023</i>, New Orleans, LA, United States, 2023.","short":"D. Xu, D. Yao, S. Lachapelle, P. Taslakian, J. von Kügelgen, F. Locatello, S. Magliacane, in:, Causal Representation Learning Workshop at NeurIPS 2023, OpenReview, 2023.","apa":"Xu, D., Yao, D., Lachapelle, S., Taslakian, P., von Kügelgen, J., Locatello, F., &#38; Magliacane, S. (2023). A sparsity principle for partially observable causal representation learning. In <i>Causal Representation Learning Workshop at NeurIPS 2023</i>. New Orleans, LA, United States: OpenReview.","chicago":"Xu, Danru, Dingling Yao, Sebastien Lachapelle, Perouz Taslakian, Julius von Kügelgen, Francesco Locatello, and Sara Magliacane. “A Sparsity Principle for Partially Observable Causal Representation Learning.” In <i>Causal Representation Learning Workshop at NeurIPS 2023</i>. OpenReview, 2023.","ista":"Xu D, Yao D, Lachapelle S, Taslakian P, von Kügelgen J, Locatello F, Magliacane S. 2023. A sparsity principle for partially observable causal representation learning. Causal Representation Learning Workshop at NeurIPS 2023. CRL: Causal Representation Learning Workshop at NeurIPS, 54."},"article_number":"54","oa_version":"Published Version","file_date_updated":"2024-02-13T08:50:53Z","conference":{"location":"New Orleans, LA, United States","name":"CRL: Causal Representation Learning Workshop at NeurIPS","start_date":"2023-12-15","end_date":"2023-12-15"}},{"oa":1,"arxiv":1,"title":"Shortcuts for causal discovery of nonlinear models by score matching","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2310.14246","open_access":"1"}],"department":[{"_id":"FrLo"}],"day":"22","language":[{"iso":"eng"}],"year":"2023","date_created":"2024-02-08T15:31:46Z","status":"public","article_processing_charge":"No","publication":"arXiv","publication_status":"submitted","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","date_published":"2023-10-22T00:00:00Z","month":"10","_id":"14961","article_number":"2310.14246","type":"preprint","abstract":[{"text":"The use of simulated data in the field of causal discovery is ubiquitous due to the scarcity of annotated real data. Recently, Reisach et al., 2021 highlighted the emergence of patterns in simulated linear data, which displays increasing marginal variance in the casual direction. As an ablation in their experiments, Montagna et al., 2023 found that similar patterns may emerge in\r\nnonlinear models for the variance of the score vector $\\nabla \\log p_{\\mathbf{X}}$, and introduced the ScoreSort algorithm. In this work, we formally define and characterize this score-sortability pattern of nonlinear additive noise models. We find that it defines a class of identifiable (bivariate) causal models overlapping with nonlinear additive noise models. We\r\ntheoretically demonstrate the advantages of ScoreSort in terms of statistical efficiency compared to prior state-of-the-art score matching-based methods and empirically show the score-sortability of the most common synthetic benchmarks in the literature. Our findings remark (1) the lack of diversity in the data as an important limitation in the evaluation of nonlinear causal discovery approaches, (2) the importance of thoroughly testing different settings within a problem class, and (3) the importance of analyzing statistical properties in\r\ncausal discovery, where research is often limited to defining identifiability conditions of the model. ","lang":"eng"}],"date_updated":"2024-02-12T10:03:33Z","citation":{"apa":"Montagna, F., Noceti, N., Rosasco, L., &#38; Locatello, F. (n.d.). Shortcuts for causal discovery of nonlinear models by score matching. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2310.14246\">https://doi.org/10.48550/arXiv.2310.14246</a>","short":"F. Montagna, N. Noceti, L. Rosasco, F. Locatello, ArXiv (n.d.).","ista":"Montagna F, Noceti N, Rosasco L, Locatello F. Shortcuts for causal discovery of nonlinear models by score matching. arXiv, 2310.14246.","chicago":"Montagna, Francesco, Nicoletta Noceti, Lorenzo Rosasco, and Francesco Locatello. “Shortcuts for Causal Discovery of Nonlinear Models by Score Matching.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2310.14246\">https://doi.org/10.48550/arXiv.2310.14246</a>.","ama":"Montagna F, Noceti N, Rosasco L, Locatello F. Shortcuts for causal discovery of nonlinear models by score matching. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2310.14246\">10.48550/arXiv.2310.14246</a>","ieee":"F. Montagna, N. Noceti, L. Rosasco, and F. Locatello, “Shortcuts for causal discovery of nonlinear models by score matching,” <i>arXiv</i>. .","mla":"Montagna, Francesco, et al. “Shortcuts for Causal Discovery of Nonlinear Models by Score Matching.” <i>ArXiv</i>, 2310.14246, doi:<a href=\"https://doi.org/10.48550/arXiv.2310.14246\">10.48550/arXiv.2310.14246</a>."},"oa_version":"Preprint","doi":"10.48550/arXiv.2310.14246","external_id":{"arxiv":["2310.14246"]},"author":[{"first_name":"Francesco","last_name":"Montagna","full_name":"Montagna, Francesco"},{"last_name":"Noceti","first_name":"Nicoletta","full_name":"Noceti, Nicoletta"},{"first_name":"Lorenzo","last_name":"Rosasco","full_name":"Rosasco, Lorenzo"},{"full_name":"Locatello, Francesco","id":"26cfd52f-2483-11ee-8040-88983bcc06d4","first_name":"Francesco","orcid":"0000-0002-4850-0683","last_name":"Locatello"}]},{"language":[{"iso":"eng"}],"year":"2023","date_created":"2024-02-08T15:33:39Z","status":"public","article_processing_charge":"No","publication":"arXiv","extern":"1","oa":1,"arxiv":1,"title":"Unsupervised open-vocabulary object localization in videos","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.2309.09858","open_access":"1"}],"department":[{"_id":"FrLo"}],"day":"18","oa_version":"Preprint","doi":"10.48550/arXiv.2309.09858","external_id":{"arxiv":["2309.09858"]},"author":[{"first_name":"Ke","last_name":"Fan","full_name":"Fan, Ke"},{"last_name":"Bai","first_name":"Zechen","full_name":"Bai, Zechen"},{"full_name":"Xiao, Tianjun","first_name":"Tianjun","last_name":"Xiao"},{"last_name":"Zietlow","first_name":"Dominik","full_name":"Zietlow, Dominik"},{"full_name":"Horn, Max","last_name":"Horn","first_name":"Max"},{"first_name":"Zixu","last_name":"Zhao","full_name":"Zhao, Zixu"},{"full_name":"Carl-Johann Simon-Gabriel, Carl-Johann Simon-Gabriel","first_name":"Carl-Johann Simon-Gabriel","last_name":"Carl-Johann Simon-Gabriel"},{"full_name":"Shou, Mike Zheng","last_name":"Shou","first_name":"Mike Zheng"},{"id":"26cfd52f-2483-11ee-8040-88983bcc06d4","orcid":"0000-0002-4850-0683","last_name":"Locatello","first_name":"Francesco","full_name":"Locatello, Francesco"},{"full_name":"Schiele, Bernt","first_name":"Bernt","last_name":"Schiele"},{"full_name":"Brox, Thomas","last_name":"Brox","first_name":"Thomas"},{"first_name":"Zheng","last_name":"Zhang","full_name":"Zhang, Zheng"},{"full_name":"Fu, Yanwei","last_name":"Fu","first_name":"Yanwei"},{"first_name":"Tong","last_name":"He","full_name":"He, Tong"}],"publication_status":"submitted","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"09","date_published":"2023-09-18T00:00:00Z","_id":"14962","article_number":"2309.09858","type":"preprint","abstract":[{"lang":"eng","text":"In this paper, we show that recent advances in video representation learning\r\nand pre-trained vision-language models allow for substantial improvements in\r\nself-supervised video object localization. We propose a method that first\r\nlocalizes objects in videos via a slot attention approach and then assigns text\r\nto the obtained slots. The latter is achieved by an unsupervised way to read\r\nlocalized semantic information from the pre-trained CLIP model. The resulting\r\nvideo object localization is entirely unsupervised apart from the implicit\r\nannotation contained in CLIP, and it is effectively the first unsupervised\r\napproach that yields good results on regular video benchmarks."}],"citation":{"mla":"Fan, Ke, et al. “Unsupervised Open-Vocabulary Object Localization in Videos.” <i>ArXiv</i>, 2309.09858, doi:<a href=\"https://doi.org/10.48550/arXiv.2309.09858\">10.48550/arXiv.2309.09858</a>.","ieee":"K. Fan <i>et al.</i>, “Unsupervised open-vocabulary object localization in videos,” <i>arXiv</i>. .","ama":"Fan K, Bai Z, Xiao T, et al. Unsupervised open-vocabulary object localization in videos. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2309.09858\">10.48550/arXiv.2309.09858</a>","chicago":"Fan, Ke, Zechen Bai, Tianjun Xiao, Dominik Zietlow, Max Horn, Zixu Zhao, Carl-Johann Simon-Gabriel Carl-Johann Simon-Gabriel, et al. “Unsupervised Open-Vocabulary Object Localization in Videos.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2309.09858\">https://doi.org/10.48550/arXiv.2309.09858</a>.","ista":"Fan K, Bai Z, Xiao T, Zietlow D, Horn M, Zhao Z, Carl-Johann Simon-Gabriel C-JS-G, Shou MZ, Locatello F, Schiele B, Brox T, Zhang Z, Fu Y, He T. Unsupervised open-vocabulary object localization in videos. arXiv, 2309.09858.","short":"K. Fan, Z. Bai, T. Xiao, D. Zietlow, M. Horn, Z. Zhao, C.-J.S.-G. Carl-Johann Simon-Gabriel, M.Z. Shou, F. Locatello, B. Schiele, T. Brox, Z. Zhang, Y. Fu, T. He, ArXiv (n.d.).","apa":"Fan, K., Bai, Z., Xiao, T., Zietlow, D., Horn, M., Zhao, Z., … He, T. (n.d.). Unsupervised open-vocabulary object localization in videos. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2309.09858\">https://doi.org/10.48550/arXiv.2309.09858</a>"},"date_updated":"2024-02-12T10:12:22Z"},{"extern":"1","publication":"arXiv","article_processing_charge":"No","status":"public","date_created":"2024-02-08T15:34:43Z","language":[{"iso":"eng"}],"year":"2023","day":"01","department":[{"_id":"FrLo"}],"main_file_link":[{"open_access":"1","url":" https://doi.org/10.48550/arXiv.2309.00233"}],"title":"Object-centric multiple object tracking","arxiv":1,"oa":1,"author":[{"full_name":"Zhao, Zixu","last_name":"Zhao","first_name":"Zixu"},{"last_name":"Wang","first_name":"Jiaze","full_name":"Wang, Jiaze"},{"first_name":"Max","last_name":"Horn","full_name":"Horn, Max"},{"full_name":"Ding, Yizhuo","first_name":"Yizhuo","last_name":"Ding"},{"full_name":"He, Tong","last_name":"He","first_name":"Tong"},{"full_name":"Bai, Zechen","first_name":"Zechen","last_name":"Bai"},{"full_name":"Zietlow, Dominik","first_name":"Dominik","last_name":"Zietlow"},{"first_name":"Carl-Johann Simon-Gabriel","last_name":"Carl-Johann Simon-Gabriel","full_name":"Carl-Johann Simon-Gabriel, Carl-Johann Simon-Gabriel"},{"first_name":"Bing","last_name":"Shuai","full_name":"Shuai, Bing"},{"full_name":"Tu, Zhuowen","first_name":"Zhuowen","last_name":"Tu"},{"full_name":"Brox, Thomas","first_name":"Thomas","last_name":"Brox"},{"full_name":"Schiele, Bernt","last_name":"Schiele","first_name":"Bernt"},{"first_name":"Yanwei","last_name":"Fu","full_name":"Fu, Yanwei"},{"full_name":"Locatello, Francesco","first_name":"Francesco","last_name":"Locatello","orcid":"0000-0002-4850-0683","id":"26cfd52f-2483-11ee-8040-88983bcc06d4"},{"full_name":"Zhang, Zheng","first_name":"Zheng","last_name":"Zhang"},{"full_name":"Xiao, Tianjun","last_name":"Xiao","first_name":"Tianjun"}],"external_id":{"arxiv":["2309.00233"]},"doi":"10.48550/arXiv.2309.00233","oa_version":"Preprint","abstract":[{"lang":"eng","text":"Unsupervised object-centric learning methods allow the partitioning of scenes\r\ninto entities without additional localization information and are excellent\r\ncandidates for reducing the annotation burden of multiple-object tracking (MOT)\r\npipelines. Unfortunately, they lack two key properties: objects are often split\r\ninto parts and are not consistently tracked over time. In fact,\r\nstate-of-the-art models achieve pixel-level accuracy and temporal consistency\r\nby relying on supervised object detection with additional ID labels for the\r\nassociation through time. This paper proposes a video object-centric model for\r\nMOT. It consists of an index-merge module that adapts the object-centric slots\r\ninto detection outputs and an object memory module that builds complete object\r\nprototypes to handle occlusions. Benefited from object-centric learning, we\r\nonly require sparse detection labels (0%-6.25%) for object localization and\r\nfeature binding. Relying on our self-supervised\r\nExpectation-Maximization-inspired loss for object association, our approach\r\nrequires no ID labels. Our experiments significantly narrow the gap between the\r\nexisting object-centric model and the fully supervised state-of-the-art and\r\noutperform several unsupervised trackers."}],"citation":{"ama":"Zhao Z, Wang J, Horn M, et al. Object-centric multiple object tracking. <i>arXiv</i>. doi:<a href=\"https://doi.org/10.48550/arXiv.2309.00233\">10.48550/arXiv.2309.00233</a>","mla":"Zhao, Zixu, et al. “Object-Centric Multiple Object Tracking.” <i>ArXiv</i>, 2309.00233, doi:<a href=\"https://doi.org/10.48550/arXiv.2309.00233\">10.48550/arXiv.2309.00233</a>.","ieee":"Z. Zhao <i>et al.</i>, “Object-centric multiple object tracking,” <i>arXiv</i>. .","short":"Z. Zhao, J. Wang, M. Horn, Y. Ding, T. He, Z. Bai, D. Zietlow, C.-J.S.-G. Carl-Johann Simon-Gabriel, B. Shuai, Z. Tu, T. Brox, B. Schiele, Y. Fu, F. Locatello, Z. Zhang, T. Xiao, ArXiv (n.d.).","apa":"Zhao, Z., Wang, J., Horn, M., Ding, Y., He, T., Bai, Z., … Xiao, T. (n.d.). Object-centric multiple object tracking. <i>arXiv</i>. <a href=\"https://doi.org/10.48550/arXiv.2309.00233\">https://doi.org/10.48550/arXiv.2309.00233</a>","chicago":"Zhao, Zixu, Jiaze Wang, Max Horn, Yizhuo Ding, Tong He, Zechen Bai, Dominik Zietlow, et al. “Object-Centric Multiple Object Tracking.” <i>ArXiv</i>, n.d. <a href=\"https://doi.org/10.48550/arXiv.2309.00233\">https://doi.org/10.48550/arXiv.2309.00233</a>.","ista":"Zhao Z, Wang J, Horn M, Ding Y, He T, Bai Z, Zietlow D, Carl-Johann Simon-Gabriel C-JS-G, Shuai B, Tu Z, Brox T, Schiele B, Fu Y, Locatello F, Zhang Z, Xiao T. Object-centric multiple object tracking. arXiv, 2309.00233."},"date_updated":"2024-02-12T10:16:21Z","type":"preprint","article_number":"2309.00233","_id":"14963","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"09","date_published":"2023-09-01T00:00:00Z","publication_status":"submitted"},{"quality_controlled":"1","oa":1,"issue":"1","day":"01","department":[{"_id":"MaIb"}],"date_created":"2024-02-14T12:12:17Z","status":"public","publisher":"Wiley","file":[{"success":1,"date_created":"2024-02-19T09:58:32Z","checksum":"7b5e8210ef1434feb173022c6dbbee0c","access_level":"open_access","file_name":"2023_InterdiscMaterials_Liu.pdf","file_id":"15015","relation":"main_file","content_type":"application/pdf","file_size":4675941,"date_updated":"2024-02-19T09:58:32Z","creator":"dernst"}],"page":"161-170","publication_status":"published","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","abstract":[{"text":"Lead sulfide (PbS) presents large potential in thermoelectric application due to its earth-abundant S element. However, its inferior average ZT (ZTave) value makes PbS less competitive with its analogs PbTe and PbSe. To promote its thermoelectric performance, this study implements strategies of continuous Se alloying and Cu interstitial doping to synergistically tune thermal and electrical transport properties in n-type PbS. First, the lattice parameter of 5.93 Å in PbS is linearly expanded to 6.03 Å in PbS0.5Se0.5 with increasing Se alloying content. This expanded lattice in Se-alloyed PbS not only intensifies phonon scattering but also facilitates the formation of Cu interstitials. Based on the PbS0.6Se0.4 content with the minimal lattice thermal conductivity, Cu interstitials are introduced to improve the electron density, thus boosting the peak power factor, from 3.88 μW cm−1 K−2 in PbS0.6Se0.4 to 20.58 μW cm−1 K−2 in PbS0.6Se0.4−1%Cu. Meanwhile, the lattice thermal conductivity in PbS0.6Se0.4−x%Cu (x = 0–2) is further suppressed due to the strong strain field caused by Cu interstitials. Finally, with the lowered thermal conductivity and high electrical transport properties, a peak ZT ~1.1 and ZTave ~0.82 can be achieved in PbS0.6Se0.4 − 1%Cu at 300–773K, which outperforms previously reported n-type PbS.","lang":"eng"}],"_id":"14985","publication_identifier":{"eissn":["2767-441X"]},"volume":2,"author":[{"full_name":"Liu, Zhengtao","last_name":"Liu","first_name":"Zhengtao"},{"full_name":"Hong, Tao","first_name":"Tao","last_name":"Hong"},{"full_name":"Xu, Liqing","first_name":"Liqing","last_name":"Xu"},{"full_name":"Wang, Sining","first_name":"Sining","last_name":"Wang"},{"full_name":"Gao, Xiang","last_name":"Gao","first_name":"Xiang"},{"first_name":"Cheng","last_name":"Chang","orcid":"0000-0002-9515-4277","id":"9E331C2E-9F27-11E9-AE48-5033E6697425","full_name":"Chang, Cheng"},{"first_name":"Xiangdong","last_name":"Ding","full_name":"Ding, Xiangdong"},{"full_name":"Xiao, Yu","last_name":"Xiao","first_name":"Yu"},{"first_name":"Li‐Dong","last_name":"Zhao","full_name":"Zhao, Li‐Dong"}],"title":"Lattice expansion enables interstitial doping to achieve a high average ZT in n‐type PbS","intvolume":"         2","acknowledgement":"The authors would like to acknowledge the strong supportof microstructure observation from Center for HighPressure Science and Technology Advanced Research(HPSTAR). We acknowledge the financial support fromthe  National  Natural  Science  Foundation  of  China:52172236, the Fundamental Research Funds for theCentral Universities: xtr042021007, Top Young TalentsProgramme of Xi'an Jiaotong University and NationalScience Fund for Distinguished Young Scholars: 51925101.","year":"2023","language":[{"iso":"eng"}],"publication":"Interdisciplinary Materials","article_processing_charge":"Yes","ddc":["540"],"date_published":"2023-01-01T00:00:00Z","month":"01","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)","image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)"},"type":"journal_article","date_updated":"2024-02-19T10:01:26Z","citation":{"short":"Z. Liu, T. Hong, L. Xu, S. Wang, X. Gao, C. Chang, X. Ding, Y. Xiao, L. Zhao, Interdisciplinary Materials 2 (2023) 161–170.","apa":"Liu, Z., Hong, T., Xu, L., Wang, S., Gao, X., Chang, C., … Zhao, L. (2023). Lattice expansion enables interstitial doping to achieve a high average ZT in n‐type PbS. <i>Interdisciplinary Materials</i>. Wiley. <a href=\"https://doi.org/10.1002/idm2.12056\">https://doi.org/10.1002/idm2.12056</a>","chicago":"Liu, Zhengtao, Tao Hong, Liqing Xu, Sining Wang, Xiang Gao, Cheng Chang, Xiangdong Ding, Yu Xiao, and Li‐Dong Zhao. “Lattice Expansion Enables Interstitial Doping to Achieve a High Average ZT in N‐type PbS.” <i>Interdisciplinary Materials</i>. Wiley, 2023. <a href=\"https://doi.org/10.1002/idm2.12056\">https://doi.org/10.1002/idm2.12056</a>.","ista":"Liu Z, Hong T, Xu L, Wang S, Gao X, Chang C, Ding X, Xiao Y, Zhao L. 2023. Lattice expansion enables interstitial doping to achieve a high average ZT in n‐type PbS. Interdisciplinary Materials. 2(1), 161–170.","ama":"Liu Z, Hong T, Xu L, et al. Lattice expansion enables interstitial doping to achieve a high average ZT in n‐type PbS. <i>Interdisciplinary Materials</i>. 2023;2(1):161-170. doi:<a href=\"https://doi.org/10.1002/idm2.12056\">10.1002/idm2.12056</a>","mla":"Liu, Zhengtao, et al. “Lattice Expansion Enables Interstitial Doping to Achieve a High Average ZT in N‐type PbS.” <i>Interdisciplinary Materials</i>, vol. 2, no. 1, Wiley, 2023, pp. 161–70, doi:<a href=\"https://doi.org/10.1002/idm2.12056\">10.1002/idm2.12056</a>.","ieee":"Z. Liu <i>et al.</i>, “Lattice expansion enables interstitial doping to achieve a high average ZT in n‐type PbS,” <i>Interdisciplinary Materials</i>, vol. 2, no. 1. Wiley, pp. 161–170, 2023."},"has_accepted_license":"1","oa_version":"Published Version","file_date_updated":"2024-02-19T09:58:32Z","doi":"10.1002/idm2.12056","article_type":"original"},{"publisher":"Internet Society","publication":"Proceedings of the 2023 Network and Distributed System Security Symposium","article_processing_charge":"No","language":[{"iso":"eng"}],"year":"2023","status":"public","date_created":"2024-02-14T14:20:40Z","acknowledgement":"This work is supported by the Novi team at Meta and funded in part by IC3 industry partners and NSF grant 1943499.","department":[{"_id":"ElKo"}],"day":"01","oa":1,"title":"Parakeet: Practical key transparency for end-to-end eEncrypted messaging","main_file_link":[{"open_access":"1","url":"https://eprint.iacr.org/2023/081"}],"quality_controlled":"1","author":[{"first_name":"Harjasleen","last_name":"Malvai","full_name":"Malvai, Harjasleen"},{"full_name":"Kokoris Kogias, Eleftherios","last_name":"Kokoris Kogias","first_name":"Eleftherios","id":"f5983044-d7ef-11ea-ac6d-fd1430a26d30"},{"full_name":"Sonnino, Alberto","last_name":"Sonnino","first_name":"Alberto"},{"first_name":"Esha","last_name":"Ghosh","full_name":"Ghosh, Esha"},{"full_name":"Oztürk, Ercan","last_name":"Oztürk","first_name":"Ercan"},{"full_name":"Lewi, Kevin","first_name":"Kevin","last_name":"Lewi"},{"first_name":"Sean","last_name":"Lawlor","full_name":"Lawlor, Sean"}],"conference":{"location":"San Diego, CA, United States","name":"NDSS: Network and Distributed Systems Security","start_date":"2023-02-27","end_date":"2023-03-03"},"doi":"10.14722/ndss.2023.24545","publication_identifier":{"isbn":["1891562835"]},"oa_version":"Published Version","_id":"14989","citation":{"ama":"Malvai H, Kokoris Kogias E, Sonnino A, et al. Parakeet: Practical key transparency for end-to-end eEncrypted messaging. In: <i>Proceedings of the 2023 Network and Distributed System Security Symposium</i>. Internet Society; 2023. doi:<a href=\"https://doi.org/10.14722/ndss.2023.24545\">10.14722/ndss.2023.24545</a>","mla":"Malvai, Harjasleen, et al. “Parakeet: Practical Key Transparency for End-to-End EEncrypted Messaging.” <i>Proceedings of the 2023 Network and Distributed System Security Symposium</i>, Internet Society, 2023, doi:<a href=\"https://doi.org/10.14722/ndss.2023.24545\">10.14722/ndss.2023.24545</a>.","ieee":"H. Malvai <i>et al.</i>, “Parakeet: Practical key transparency for end-to-end eEncrypted messaging,” in <i>Proceedings of the 2023 Network and Distributed System Security Symposium</i>, San Diego, CA, United States, 2023.","short":"H. Malvai, E. Kokoris Kogias, A. Sonnino, E. Ghosh, E. Oztürk, K. Lewi, S. Lawlor, in:, Proceedings of the 2023 Network and Distributed System Security Symposium, Internet Society, 2023.","apa":"Malvai, H., Kokoris Kogias, E., Sonnino, A., Ghosh, E., Oztürk, E., Lewi, K., &#38; Lawlor, S. (2023). Parakeet: Practical key transparency for end-to-end eEncrypted messaging. In <i>Proceedings of the 2023 Network and Distributed System Security Symposium</i>. San Diego, CA, United States: Internet Society. <a href=\"https://doi.org/10.14722/ndss.2023.24545\">https://doi.org/10.14722/ndss.2023.24545</a>","chicago":"Malvai, Harjasleen, Eleftherios Kokoris Kogias, Alberto Sonnino, Esha Ghosh, Ercan Oztürk, Kevin Lewi, and Sean Lawlor. “Parakeet: Practical Key Transparency for End-to-End EEncrypted Messaging.” In <i>Proceedings of the 2023 Network and Distributed System Security Symposium</i>. Internet Society, 2023. <a href=\"https://doi.org/10.14722/ndss.2023.24545\">https://doi.org/10.14722/ndss.2023.24545</a>.","ista":"Malvai H, Kokoris Kogias E, Sonnino A, Ghosh E, Oztürk E, Lewi K, Lawlor S. 2023. Parakeet: Practical key transparency for end-to-end eEncrypted messaging. Proceedings of the 2023 Network and Distributed System Security Symposium. NDSS: Network and Distributed Systems Security."},"date_updated":"2024-02-19T12:11:15Z","abstract":[{"text":"Encryption alone is not enough for secure end-to end encrypted messaging: a server must also honestly serve public keys to users. Key transparency has been presented as an efficient\r\nsolution for detecting (and hence deterring) a server that attempts to dishonestly serve keys. Key transparency involves two major components: (1) a username to public key mapping, stored and cryptographically committed to by the server, and, (2) an outof-band consistency protocol for serving short commitments to users. In the setting of real-world deployments and supporting production scale, new challenges must be considered for both of these components. We enumerate these challenges and provide solutions to address them. In particular, we design and implement a memory-optimized and privacy-preserving verifiable data structure for committing to the username to public key store.\r\nTo make this implementation viable for production, we also integrate support for persistent and distributed storage. We also propose a future-facing solution, termed “compaction”, as\r\na mechanism for mitigating practical issues that arise from dealing with infinitely growing server data structures. Finally, we implement a consensusless solution that achieves the minimum requirements for a service that consistently distributes commitments for a transparency application, providing a much more efficient protocol for distributing small and consistent\r\ncommitments to users. This culminates in our production-grade implementation of a key transparency system (Parakeet) which we have open-sourced, along with a demonstration of feasibility through our benchmarks.","lang":"eng"}],"type":"conference","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","date_published":"2023-03-01T00:00:00Z","month":"03","publication_status":"published"},{"doi":"10.5281/ZENODO.7548214","author":[{"last_name":"Meggendorfer","orcid":"0000-0002-1712-2165","first_name":"Tobias","id":"b21b0c15-30a2-11eb-80dc-f13ca25802e1","full_name":"Meggendorfer, Tobias"}],"related_material":{"record":[{"status":"public","relation":"used_in_publication","id":"13139"}]},"oa_version":"Published Version","type":"research_data_reference","date_updated":"2024-02-27T07:19:32Z","has_accepted_license":"1","citation":{"apa":"Meggendorfer, T. (2023). Artefact for: Correct Approximation of Stationary Distributions. Zenodo. <a href=\"https://doi.org/10.5281/ZENODO.7548214\">https://doi.org/10.5281/ZENODO.7548214</a>","short":"T. Meggendorfer, (2023).","ista":"Meggendorfer T. 2023. Artefact for: Correct Approximation of Stationary Distributions, Zenodo, <a href=\"https://doi.org/10.5281/ZENODO.7548214\">10.5281/ZENODO.7548214</a>.","chicago":"Meggendorfer, Tobias. “Artefact for: Correct Approximation of Stationary Distributions.” Zenodo, 2023. <a href=\"https://doi.org/10.5281/ZENODO.7548214\">https://doi.org/10.5281/ZENODO.7548214</a>.","ama":"Meggendorfer T. Artefact for: Correct Approximation of Stationary Distributions. 2023. doi:<a href=\"https://doi.org/10.5281/ZENODO.7548214\">10.5281/ZENODO.7548214</a>","ieee":"T. Meggendorfer, “Artefact for: Correct Approximation of Stationary Distributions.” Zenodo, 2023.","mla":"Meggendorfer, Tobias. <i>Artefact for: Correct Approximation of Stationary Distributions</i>. Zenodo, 2023, doi:<a href=\"https://doi.org/10.5281/ZENODO.7548214\">10.5281/ZENODO.7548214</a>."},"abstract":[{"text":"The software artefact to evaluate the approximation of stationary distributions implementation.","lang":"eng"}],"_id":"14990","ddc":["000"],"user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"01","date_published":"2023-01-18T00:00:00Z","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)","image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)"},"article_processing_charge":"No","publisher":"Zenodo","date_created":"2024-02-14T14:27:06Z","status":"public","year":"2023","day":"18","department":[{"_id":"KrCh"}],"main_file_link":[{"url":"https://doi.org/10.5281/zenodo.7548214","open_access":"1"}],"title":"Artefact for: Correct Approximation of Stationary Distributions","oa":1},{"_id":"14991","type":"research_data_reference","abstract":[{"text":"This repository contains the data, scripts, WRF codes and files required to reproduce the results of the manuscript \"Assessing Memory in Convection Schemes Using Idealized Tests\" submitted to the Journal of Advances in Modeling Earth Systems (JAMES).","lang":"eng"}],"has_accepted_license":"1","citation":{"ama":"Hwong Y-L, Colin M, Aglas P, Muller CJ, Sherwood SC. Data-assessing memory in convection schemes using idealized tests. 2023. doi:<a href=\"https://doi.org/10.5281/ZENODO.7757041\">10.5281/ZENODO.7757041</a>","mla":"Hwong, Yi-Ling, et al. <i>Data-Assessing Memory in Convection Schemes Using Idealized Tests</i>. Zenodo, 2023, doi:<a href=\"https://doi.org/10.5281/ZENODO.7757041\">10.5281/ZENODO.7757041</a>.","ieee":"Y.-L. Hwong, M. Colin, P. Aglas, C. J. Muller, and S. C. Sherwood, “Data-assessing memory in convection schemes using idealized tests.” Zenodo, 2023.","short":"Y.-L. Hwong, M. Colin, P. Aglas, C.J. Muller, S.C. Sherwood, (2023).","apa":"Hwong, Y.-L., Colin, M., Aglas, P., Muller, C. J., &#38; Sherwood, S. C. (2023). Data-assessing memory in convection schemes using idealized tests. Zenodo. <a href=\"https://doi.org/10.5281/ZENODO.7757041\">https://doi.org/10.5281/ZENODO.7757041</a>","chicago":"Hwong, Yi-Ling, Maxime Colin, Philipp Aglas, Caroline J Muller, and Steven C. Sherwood. “Data-Assessing Memory in Convection Schemes Using Idealized Tests.” Zenodo, 2023. <a href=\"https://doi.org/10.5281/ZENODO.7757041\">https://doi.org/10.5281/ZENODO.7757041</a>.","ista":"Hwong Y-L, Colin M, Aglas P, Muller CJ, Sherwood SC. 2023. Data-assessing memory in convection schemes using idealized tests, Zenodo, <a href=\"https://doi.org/10.5281/ZENODO.7757041\">10.5281/ZENODO.7757041</a>."},"date_updated":"2024-02-27T07:26:31Z","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)","image":"/images/cc_by.png","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)"},"ddc":["550"],"date_published":"2023-06-23T00:00:00Z","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"06","ec_funded":1,"doi":"10.5281/ZENODO.7757041","author":[{"orcid":"0000-0001-9281-3479","last_name":"Hwong","first_name":"Yi-Ling","id":"1217aa61-4dd1-11ec-9ac3-f2ba3f17ee22","full_name":"Hwong, Yi-Ling"},{"last_name":"Colin","first_name":"Maxime","full_name":"Colin, Maxime"},{"id":"02eace56-97fc-11ee-b81a-f0939ca85a77","last_name":"Aglas","first_name":"Philipp","full_name":"Aglas, Philipp"},{"full_name":"Muller, Caroline J","id":"f978ccb0-3f7f-11eb-b193-b0e2bd13182b","first_name":"Caroline J","orcid":"0000-0001-5836-5350","last_name":"Muller"},{"last_name":"Sherwood","first_name":"Steven C.","full_name":"Sherwood, Steven C."}],"related_material":{"record":[{"status":"public","id":"14654","relation":"used_in_publication"}]},"project":[{"_id":"fc2ed2f7-9c52-11eb-aca3-c01059dda49c","name":"IST-BRIDGE: International postdoctoral program","grant_number":"101034413","call_identifier":"H2020"}],"oa_version":"Published Version","department":[{"_id":"CaMu"}],"day":"23","oa":1,"title":"Data-assessing memory in convection schemes using idealized tests","main_file_link":[{"url":"https://doi.org/10.5281/zenodo.7757041","open_access":"1"}],"article_processing_charge":"No","publisher":"Zenodo","year":"2023","date_created":"2024-02-14T14:37:57Z","status":"public"},{"abstract":[{"text":"In this chapter we first review the Levy–Lieb functional, which gives the lowest kinetic and interaction energy that can be reached with all possible quantum states having a given density. We discuss two possible convex generalizations of this functional, corresponding to using mixed canonical and grand-canonical states, respectively. We present some recent works about the local density approximation, in which the functionals get replaced by purely local functionals constructed using the uniform electron gas energy per unit volume. We then review the known upper and lower bounds on the Levy–Lieb functionals. We start with the kinetic energy alone, then turn to the classical interaction alone, before we are able to put everything together. A later section is devoted to the Hohenberg–Kohn theorem and the role of many-body unique continuation in its proof.","lang":"eng"}],"edition":"1","_id":"14992","publication_status":"published","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","author":[{"full_name":"Lewin, Mathieu","first_name":"Mathieu","last_name":"Lewin"},{"first_name":"Elliott H.","last_name":"Lieb","full_name":"Lieb, Elliott H."},{"full_name":"Seiringer, Robert","id":"4AFD0470-F248-11E8-B48F-1D18A9856A87","first_name":"Robert","orcid":"0000-0002-6781-0521","last_name":"Seiringer"}],"external_id":{"arxiv":["1912.10424"]},"publication_identifier":{"isbn":["9783031223396"],"eisbn":["9783031223402"],"issn":["3005-0286"]},"day":"19","series_title":"MAMOMO","department":[{"_id":"RoSe"}],"quality_controlled":"1","oa":1,"arxiv":1,"page":"115-182","publisher":"Springer","date_created":"2024-02-14T14:44:33Z","status":"public","type":"book_chapter","citation":{"ieee":"M. Lewin, E. H. Lieb, and R. Seiringer, “Universal Functionals in Density Functional Theory,” in <i>Density Functional Theory</i>, 1st ed., E. Cances and G. Friesecke, Eds. Springer, 2023, pp. 115–182.","mla":"Lewin, Mathieu, et al. “Universal Functionals in Density Functional Theory.” <i>Density Functional Theory</i>, edited by Eric Cances and Gero Friesecke, 1st ed., Springer, 2023, pp. 115–82, doi:<a href=\"https://doi.org/10.1007/978-3-031-22340-2_3\">10.1007/978-3-031-22340-2_3</a>.","ama":"Lewin M, Lieb EH, Seiringer R. Universal Functionals in Density Functional Theory. In: Cances E, Friesecke G, eds. <i>Density Functional Theory</i>. 1st ed. MAMOMO. Springer; 2023:115-182. doi:<a href=\"https://doi.org/10.1007/978-3-031-22340-2_3\">10.1007/978-3-031-22340-2_3</a>","ista":"Lewin M, Lieb EH, Seiringer R. 2023.Universal Functionals in Density Functional Theory. In: Density Functional Theory. Mathematics and Molecular Modeling, , 115–182.","chicago":"Lewin, Mathieu, Elliott H. Lieb, and Robert Seiringer. “Universal Functionals in Density Functional Theory.” In <i>Density Functional Theory</i>, edited by Eric Cances and Gero Friesecke, 1st ed., 115–82. MAMOMO. Springer, 2023. <a href=\"https://doi.org/10.1007/978-3-031-22340-2_3\">https://doi.org/10.1007/978-3-031-22340-2_3</a>.","apa":"Lewin, M., Lieb, E. H., &#38; Seiringer, R. (2023). Universal Functionals in Density Functional Theory. In E. Cances &#38; G. Friesecke (Eds.), <i>Density Functional Theory</i> (1st ed., pp. 115–182). Springer. <a href=\"https://doi.org/10.1007/978-3-031-22340-2_3\">https://doi.org/10.1007/978-3-031-22340-2_3</a>","short":"M. Lewin, E.H. Lieb, R. Seiringer, in:, E. Cances, G. Friesecke (Eds.), Density Functional Theory, 1st ed., Springer, 2023, pp. 115–182."},"date_updated":"2024-02-20T08:33:06Z","date_published":"2023-07-19T00:00:00Z","month":"07","editor":[{"full_name":"Cances, Eric","first_name":"Eric","last_name":"Cances"},{"first_name":"Gero","last_name":"Friesecke","full_name":"Friesecke, Gero"}],"doi":"10.1007/978-3-031-22340-2_3","alternative_title":["Mathematics and Molecular Modeling"],"oa_version":"Preprint","main_file_link":[{"url":"https://doi.org/10.48550/arXiv.1912.10424","open_access":"1"}],"title":"Universal Functionals in Density Functional Theory","article_processing_charge":"No","publication":"Density Functional Theory","year":"2023","language":[{"iso":"eng"}]}]