[{"degree_awarded":"PhD","ec_funded":1,"status":"public","abstract":[{"lang":"eng","text":"Plant motions occur across a wide spectrum of timescales, ranging from seed dispersal through bursting (milliseconds) and stomatal opening (minutes) to long-term adaptation of gross architecture. Relatively fast motions include water-driven growth as exemplified by root cell expansion under abiotic/biotic stresses or during gravitropism. A showcase is a root growth inhibition in 30 seconds triggered by the phytohormone auxin. However, the cellular and molecular mechanisms are still largely unknown. This thesis covers the studies about this topic as follows. By taking advantage of microfluidics combined with live imaging, pharmaceutical tools, and transgenic lines, we examined the kinetics of and causal relationship among various auxininduced rapid cellular changes in root growth, apoplastic pH, cytosolic Ca2+, cortical microtubule (CMT) orientation, and vacuolar morphology. We revealed that CMT reorientation and vacuolar constriction are the consequence of growth itself instead of responding directly to auxin. In contrast, auxin induces apoplast alkalinization to rapidly inhibit root growth in 30 seconds. This auxin-triggered apoplast alkalinization results from rapid H+- influx that is contributed by Ca2+ inward channel CYCLIC NUCLEOTIDE-GATED CHANNEL 14 (CNGC14)-dependent Ca2+ signaling. To dissect which auxin signaling mediates the rapid apoplast alkalinization, we\r\ncombined microfluidics and genetic engineering to verify that TIR1/AFB receptors conduct a non-transcriptional regulation on Ca2+ and H+ -influx. This non-canonical pathway is mostly mediated by the cytosolic portion of TIR1/AFB. On the other hand, we uncovered, using biochemical and phospho-proteomic analysis, that auxin cell surface signaling component TRANSMEMBRANE KINASE 1 (TMK1) plays a negative role during auxin-trigger apoplast\r\nalkalinization and root growth inhibition through directly activating PM H+ -ATPases. Therefore, we discovered that PM H+ -ATPases counteract instead of mediate the auxintriggered rapid H+ -influx, and that TIR1/AFB and TMK1 regulate root growth antagonistically. This opposite effect of TIR1/AFB and TMK1 is consistent during auxin-induced hypocotyl elongation, leading us to explore the relation of two signaling pathways. Assisted with biochemistry and fluorescent imaging, we verified for the first time that TIR1/AFB and TMK1 can interact with each other. The ability of TIR1/AFB binding to membrane lipid provides a basis for the interaction of plasma membrane- and cytosol-localized proteins.\r\nBesides, transgenic analysis combined with genetic engineering and biochemistry showed that  vi\r\nthey do function in the same pathway. Particularly, auxin-induced TMK1 increase is TIR1/AFB dependent, suggesting TIR1/AFB regulation on TMK1. Conversely, TMK1 also regulates TIR1/AFB protein levels and thus auxin canonical signaling. To follow the study of rapid growth regulation, we analyzed another rapid growth regulator, signaling peptide RALF1. We showed that RALF1 also triggers a rapid and reversible growth inhibition caused by H + influx, highly resembling but not dependent on auxin. Besides, RALF1 promotes auxin biosynthesis by increasing expression of auxin biosynthesis enzyme YUCCAs and thus induces auxin signaling in ca. 1 hour, contributing to the sustained RALF1-triggered growth inhibition. These studies collectively contribute to understanding rapid regulation on plant cell\r\ngrowth, novel auxin signaling pathway as well as auxin-peptide crosstalk. "}],"day":"06","has_accepted_license":"1","supervisor":[{"last_name":"Friml","first_name":"Jiří","full_name":"Friml, Jiří","id":"4159519E-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-8302-7596"}],"citation":{"ama":"Li L. Rapid cell growth regulation in Arabidopsis. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10083\">10.15479/at:ista:10083</a>","short":"L. Li, Rapid Cell Growth Regulation in Arabidopsis, Institute of Science and Technology Austria, 2021.","ieee":"L. Li, “Rapid cell growth regulation in Arabidopsis,” Institute of Science and Technology Austria, 2021.","mla":"Li, Lanxin. <i>Rapid Cell Growth Regulation in Arabidopsis</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10083\">10.15479/at:ista:10083</a>.","apa":"Li, L. (2021). <i>Rapid cell growth regulation in Arabidopsis</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10083\">https://doi.org/10.15479/at:ista:10083</a>","ista":"Li L. 2021. Rapid cell growth regulation in Arabidopsis. Institute of Science and Technology Austria.","chicago":"Li, Lanxin. “Rapid Cell Growth Regulation in Arabidopsis.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10083\">https://doi.org/10.15479/at:ista:10083</a>."},"user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","_id":"10083","month":"10","title":"Rapid cell growth regulation in Arabidopsis","related_material":{"record":[{"status":"public","relation":"part_of_dissertation","id":"442"},{"relation":"part_of_dissertation","status":"public","id":"8931"},{"relation":"part_of_dissertation","status":"public","id":"9287"},{"id":"8283","status":"public","relation":"part_of_dissertation"},{"relation":"part_of_dissertation","status":"public","id":"8986"},{"id":"10015","relation":"part_of_dissertation","status":"public"},{"id":"10095","status":"public","relation":"part_of_dissertation"},{"id":"6627","relation":"part_of_dissertation","status":"public"}]},"oa":1,"publication_status":"published","department":[{"_id":"GradSch"},{"_id":"JiFr"}],"alternative_title":["ISTA Thesis"],"date_created":"2021-10-04T13:33:10Z","year":"2021","date_updated":"2025-05-07T11:12:33Z","project":[{"_id":"2564DBCA-B435-11E9-9278-68D0E5697425","name":"International IST Doctoral Program","grant_number":"665385","call_identifier":"H2020"},{"name":"A Case Study of Plant Growth Regulation: Molecular Mechanism of Auxin-mediated Rapid Growth Inhibition in Arabidopsis Root","grant_number":"25351","_id":"26B4D67E-B435-11E9-9278-68D0E5697425"}],"file":[{"relation":"main_file","checksum":"3b2f55b3b8ae05337a0dcc1cd8595b10","file_size":8616142,"creator":"cchlebak","file_name":"0._IST_Austria_Thesis_Lanxin_Li_1014_pdftron.pdf","embargo":"2022-10-14","content_type":"application/pdf","date_created":"2021-10-14T08:00:07Z","file_id":"10138","date_updated":"2022-12-20T23:30:03Z","access_level":"open_access"},{"file_name":"0._IST_Austria_Thesis_Lanxin_Li_1014.docx","access_level":"closed","date_updated":"2022-12-20T23:30:03Z","embargo_to":"open_access","content_type":"application/vnd.openxmlformats-officedocument.wordprocessingml.document","date_created":"2021-10-14T08:00:13Z","file_id":"10139","checksum":"f23ed258ca894f6aabf58b0c128bf242","relation":"source_file","creator":"cchlebak","file_size":15058499}],"article_processing_charge":"No","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by-nc-nd/4.0/legalcode","name":"Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)","image":"/images/cc_by_nc_nd.png","short":"CC BY-NC-ND (4.0)"},"publication_identifier":{"issn":["2663-337X"]},"doi":"10.15479/at:ista:10083","publisher":"Institute of Science and Technology Austria","language":[{"iso":"eng"}],"file_date_updated":"2022-12-20T23:30:03Z","type":"dissertation","oa_version":"Published Version","date_published":"2021-10-06T00:00:00Z","ddc":["575"],"author":[{"first_name":"Lanxin","full_name":"Li, Lanxin","last_name":"Li"}]},{"type":"dissertation","file_date_updated":"2022-12-20T23:30:05Z","language":[{"iso":"eng"}],"publisher":"Institute of Science and Technology Austria","doi":"10.15479/at:ista:10135","author":[{"first_name":"Hana","full_name":"Semerádová, Hana","last_name":"Semerádová","id":"42FE702E-F248-11E8-B48F-1D18A9856A87"}],"ddc":["570"],"date_published":"2021-10-13T00:00:00Z","oa_version":"Published Version","file":[{"date_updated":"2022-12-20T23:30:05Z","access_level":"closed","date_created":"2021-10-27T07:45:37Z","embargo_to":"open_access","file_id":"10186","content_type":"application/vnd.openxmlformats-officedocument.wordprocessingml.document","file_name":"Hana_Semeradova_Disertation_Thesis_II_Revised_3.docx","creator":"cziletti","file_size":28508629,"checksum":"ce7108853e6cec6224f17cd6429b51fe","relation":"source_file"},{"creator":"cziletti","file_size":10623525,"checksum":"0d7afb846e8e31ec794de47bf44e12ef","relation":"main_file","date_updated":"2022-12-20T23:30:05Z","access_level":"open_access","content_type":"application/pdf","date_created":"2021-10-27T07:45:57Z","file_id":"10187","embargo":"2022-10-28","file_name":"Hana_Semeradova_Disertation_Thesis_II_Revised_3PDFA.pdf"}],"project":[{"grant_number":"24746","name":"Molecular mechanisms of the cytokinin regulated endomembrane trafficking to coordinate plant organogenesis.","_id":"261821BC-B435-11E9-9278-68D0E5697425"}],"date_updated":"2024-01-25T10:53:29Z","year":"2021","date_created":"2021-10-13T13:42:48Z","publication_identifier":{"isbn":["978-3-99078-014-5"],"issn":["2663-337X"]},"article_processing_charge":"No","publication_status":"published","oa":1,"title":"Molecular mechanisms of the cytokinin-regulated endomembrane trafficking to coordinate plant organogenesis","related_material":{"record":[{"id":"9160","relation":"part_of_dissertation","status":"public"}]},"_id":"10135","month":"10","user_id":"8b945eb4-e2f2-11eb-945a-df72226e66a9","alternative_title":["ISTA Thesis"],"department":[{"_id":"GradSch"},{"_id":"EvBe"}],"abstract":[{"lang":"eng","text":"Plants maintain the capacity to develop new organs e.g. lateral roots post-embryonically throughout their whole life and thereby flexibly adapt to ever-changing environmental conditions. Plant hormones auxin and cytokinin are the main regulators of the lateral root organogenesis. Additionally to their solo activities, the interaction between auxin and\r\ncytokinin plays crucial role in fine-tuning of lateral root development and growth. In particular, cytokinin modulates auxin distribution within the developing lateral root by affecting the endomembrane trafficking of auxin transporter PIN1 and promoting its vacuolar degradation (Marhavý et al., 2011, 2014). This effect is independent of transcription and\r\ntranslation. Therefore, it suggests novel, non-canonical cytokinin activity occuring possibly on the posttranslational level. Impact of cytokinin and other plant hormones on auxin transporters (including PIN1) on the posttranslational level is described in detail in the introduction part of this thesis in a form of a review (Semeradova et al., 2020). To gain insights into the molecular machinery underlying cytokinin effect on the endomembrane trafficking in the plant cell, in particular on the PIN1 degradation, we conducted two large proteomic screens: 1) Identification of cytokinin binding proteins using\r\nchemical proteomics. 2) Monitoring of proteomic and phosphoproteomic changes upon cytokinin treatment. In the first screen, we identified DYNAMIN RELATED PROTEIN 2A (DRP2A). We found that DRP2A plays a role in cytokinin regulated processes during the plant growth and that cytokinin treatment promotes destabilization of DRP2A protein. However, the role of DRP2A in the PIN1 degradation remains to be elucidated. In the second screen, we found VACUOLAR PROTEIN SORTING 9A (VPS9A). VPS9a plays crucial role in plant’s response to cytokin and in cytokinin mediated PIN1 degradation. Altogether, we identified proteins, which bind to cytokinin and proteins that in response to\r\ncytokinin exhibit significantly changed abundance or phosphorylation pattern. By combining information from these two screens, we can pave our way towards understanding of noncanonical cytokinin effects."}],"status":"public","degree_awarded":"PhD","citation":{"ama":"Semerádová H. Molecular mechanisms of the cytokinin-regulated endomembrane trafficking to coordinate plant organogenesis. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10135\">10.15479/at:ista:10135</a>","short":"H. Semerádová, Molecular Mechanisms of the Cytokinin-Regulated Endomembrane Trafficking to Coordinate Plant Organogenesis, Institute of Science and Technology Austria, 2021.","ieee":"H. Semerádová, “Molecular mechanisms of the cytokinin-regulated endomembrane trafficking to coordinate plant organogenesis,” Institute of Science and Technology Austria, 2021.","mla":"Semerádová, Hana. <i>Molecular Mechanisms of the Cytokinin-Regulated Endomembrane Trafficking to Coordinate Plant Organogenesis</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10135\">10.15479/at:ista:10135</a>.","chicago":"Semerádová, Hana. “Molecular Mechanisms of the Cytokinin-Regulated Endomembrane Trafficking to Coordinate Plant Organogenesis.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10135\">https://doi.org/10.15479/at:ista:10135</a>.","ista":"Semerádová H. 2021. Molecular mechanisms of the cytokinin-regulated endomembrane trafficking to coordinate plant organogenesis. Institute of Science and Technology Austria.","apa":"Semerádová, H. (2021). <i>Molecular mechanisms of the cytokinin-regulated endomembrane trafficking to coordinate plant organogenesis</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10135\">https://doi.org/10.15479/at:ista:10135</a>"},"supervisor":[{"id":"38F4F166-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-8510-9739","full_name":"Benková, Eva","first_name":"Eva","last_name":"Benková"}],"has_accepted_license":"1","day":"13"},{"publication_identifier":{"eissn":["2475-1421"]},"scopus_import":"1","article_processing_charge":"No","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)","image":"/images/cc_by.png","short":"CC BY (4.0)"},"volume":5,"file":[{"content_type":"application/pdf","date_created":"2021-11-04T07:24:48Z","file_id":"10215","success":1,"access_level":"open_access","date_updated":"2021-11-04T07:24:48Z","file_name":"2021_ProcACMPL_Bui.pdf","file_size":2903485,"creator":"cchlebak","relation":"main_file","checksum":"9d6dce7b611853c529bb7b1915ac579e"}],"quality_controlled":"1","date_created":"2021-10-27T15:05:34Z","year":"2021","date_updated":"2025-07-14T09:10:16Z","project":[{"name":"Formal Methods for Stochastic Models: Algorithms and Applications","grant_number":"863818","_id":"0599E47C-7A3F-11EA-A408-12923DDC885E","call_identifier":"H2020"},{"_id":"25892FC0-B435-11E9-9278-68D0E5697425","grant_number":"ICT15-003","name":"Efficient Algorithms for Computer Aided Verification"}],"issue":"OOPSLA","publication":"Proceedings of the ACM on Programming Languages","ddc":["000"],"author":[{"first_name":"Truc Lam","full_name":"Bui, Truc Lam","last_name":"Bui"},{"orcid":"0000-0002-4561-241X","id":"2E5DCA20-F248-11E8-B48F-1D18A9856A87","last_name":"Chatterjee","first_name":"Krishnendu","full_name":"Chatterjee, Krishnendu"},{"last_name":"Gautam","first_name":"Tushar","full_name":"Gautam, Tushar"},{"id":"49704004-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-8943-0722","last_name":"Pavlogiannis","first_name":"Andreas","full_name":"Pavlogiannis, Andreas"},{"orcid":"0000-0001-9036-063X","id":"3AF3DA7C-F248-11E8-B48F-1D18A9856A87","last_name":"Toman","full_name":"Toman, Viktor","first_name":"Viktor"}],"external_id":{"arxiv":["2011.11763"]},"oa_version":"Published Version","date_published":"2021-10-15T00:00:00Z","arxiv":1,"language":[{"iso":"eng"}],"file_date_updated":"2021-11-04T07:24:48Z","type":"journal_article","doi":"10.1145/3485541","publisher":"Association for Computing Machinery","keyword":["safety","risk","reliability and quality","software"],"citation":{"short":"T.L. Bui, K. Chatterjee, T. Gautam, A. Pavlogiannis, V. Toman, Proceedings of the ACM on Programming Languages 5 (2021).","ama":"Bui TL, Chatterjee K, Gautam T, Pavlogiannis A, Toman V. The reads-from equivalence for the TSO and PSO memory models. <i>Proceedings of the ACM on Programming Languages</i>. 2021;5(OOPSLA). doi:<a href=\"https://doi.org/10.1145/3485541\">10.1145/3485541</a>","chicago":"Bui, Truc Lam, Krishnendu Chatterjee, Tushar Gautam, Andreas Pavlogiannis, and Viktor Toman. “The Reads-from Equivalence for the TSO and PSO Memory Models.” <i>Proceedings of the ACM on Programming Languages</i>. Association for Computing Machinery, 2021. <a href=\"https://doi.org/10.1145/3485541\">https://doi.org/10.1145/3485541</a>.","ista":"Bui TL, Chatterjee K, Gautam T, Pavlogiannis A, Toman V. 2021. The reads-from equivalence for the TSO and PSO memory models. Proceedings of the ACM on Programming Languages. 5(OOPSLA), 164.","apa":"Bui, T. L., Chatterjee, K., Gautam, T., Pavlogiannis, A., &#38; Toman, V. (2021). The reads-from equivalence for the TSO and PSO memory models. <i>Proceedings of the ACM on Programming Languages</i>. Association for Computing Machinery. <a href=\"https://doi.org/10.1145/3485541\">https://doi.org/10.1145/3485541</a>","mla":"Bui, Truc Lam, et al. “The Reads-from Equivalence for the TSO and PSO Memory Models.” <i>Proceedings of the ACM on Programming Languages</i>, vol. 5, no. OOPSLA, 164, Association for Computing Machinery, 2021, doi:<a href=\"https://doi.org/10.1145/3485541\">10.1145/3485541</a>.","ieee":"T. L. Bui, K. Chatterjee, T. Gautam, A. Pavlogiannis, and V. Toman, “The reads-from equivalence for the TSO and PSO memory models,” <i>Proceedings of the ACM on Programming Languages</i>, vol. 5, no. OOPSLA. Association for Computing Machinery, 2021."},"intvolume":"         5","day":"15","has_accepted_license":"1","article_number":"164","status":"public","abstract":[{"text":"In this work we solve the algorithmic problem of consistency verification for the TSO and PSO memory models given a reads-from map, denoted VTSO-rf and VPSO-rf, respectively. For an execution of n events over k threads and d variables, we establish novel bounds that scale as nk+1 for TSO and as nk+1· min(nk2, 2k· d) for PSO. Moreover, based on our solution to these problems, we develop an SMC algorithm under TSO and PSO that uses the RF equivalence. The algorithm is exploration-optimal, in the sense that it is guaranteed to explore each class of the RF partitioning exactly once, and spends polynomial time per class when k is bounded. Finally, we implement all our algorithms in the SMC tool Nidhugg, and perform a large number of experiments over benchmarks from existing literature. Our experimental results show that our algorithms for VTSO-rf and VPSO-rf provide significant scalability improvements over standard alternatives. Moreover, when used for SMC, the RF partitioning is often much coarser than the standard Shasha-Snir partitioning for TSO/PSO, which yields a significant speedup in the model checking task.\r\n\r\n","lang":"eng"}],"ec_funded":1,"acknowledgement":"The research was partially funded by the ERC CoG 863818 (ForM-SMArt) and the Vienna Science\r\nand Technology Fund (WWTF) through project ICT15-003.","department":[{"_id":"GradSch"},{"_id":"KrCh"}],"title":"The reads-from equivalence for the TSO and PSO memory models","related_material":{"record":[{"id":"10199","status":"public","relation":"dissertation_contains"}]},"oa":1,"publication_status":"published","article_type":"original","user_id":"8b945eb4-e2f2-11eb-945a-df72226e66a9","month":"10","_id":"10191"},{"citation":{"ama":"Toman V. Improved verification techniques for concurrent systems. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10199\">10.15479/at:ista:10199</a>","short":"V. Toman, Improved Verification Techniques for Concurrent Systems, Institute of Science and Technology Austria, 2021.","ieee":"V. Toman, “Improved verification techniques for concurrent systems,” Institute of Science and Technology Austria, 2021.","mla":"Toman, Viktor. <i>Improved Verification Techniques for Concurrent Systems</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10199\">10.15479/at:ista:10199</a>.","apa":"Toman, V. (2021). <i>Improved verification techniques for concurrent systems</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10199\">https://doi.org/10.15479/at:ista:10199</a>","ista":"Toman V. 2021. Improved verification techniques for concurrent systems. Institute of Science and Technology Austria.","chicago":"Toman, Viktor. “Improved Verification Techniques for Concurrent Systems.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10199\">https://doi.org/10.15479/at:ista:10199</a>."},"keyword":["concurrency","verification","model checking"],"supervisor":[{"last_name":"Chatterjee","first_name":"Krishnendu","full_name":"Chatterjee, Krishnendu","id":"2E5DCA20-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-4561-241X"}],"has_accepted_license":"1","day":"31","abstract":[{"lang":"eng","text":"The design and verification of concurrent systems remains an open challenge due to the non-determinism that arises from the inter-process communication. In particular, concurrent programs are notoriously difficult both to be written correctly and to be analyzed formally, as complex thread interaction has to be accounted for. The difficulties are further exacerbated when concurrent programs get executed on modern-day hardware, which contains various buffering and caching mechanisms for efficiency reasons. This causes further subtle non-determinism, which can often produce very unintuitive behavior of the concurrent programs. Model checking is at the forefront of tackling the verification problem, where the task is to decide, given as input a concurrent system and a desired property, whether the system satisfies the property. The inherent state-space explosion problem in model checking of concurrent systems causes naïve explicit methods not to scale, thus more inventive methods are required. One such method is stateless model checking (SMC), which explores in memory-efficient manner the program executions rather than the states of the program. State-of-the-art SMC is typically coupled with partial order reduction (POR) techniques, which argue that certain executions provably produce identical system behavior, thus limiting the amount of executions one needs to explore in order to cover all possible behaviors. Another method to tackle the state-space explosion is symbolic model checking, where the considered techniques operate on a succinct implicit representation of the input system rather than explicitly accessing the system. In this thesis we present new techniques for verification of concurrent systems. We present several novel POR methods for SMC of concurrent programs under various models of semantics, some of which account for write-buffering mechanisms. Additionally, we present novel algorithms for symbolic model checking of finite-state concurrent systems, where the desired property of the systems is to ensure a formally defined notion of fairness."}],"status":"public","ec_funded":1,"degree_awarded":"PhD","alternative_title":["ISTA Thesis"],"department":[{"_id":"GradSch"},{"_id":"KrCh"}],"publication_status":"published","related_material":{"record":[{"id":"10190","status":"public","relation":"part_of_dissertation"},{"id":"9987","status":"public","relation":"part_of_dissertation"},{"id":"141","status":"public","relation":"part_of_dissertation"},{"id":"10191","relation":"part_of_dissertation","status":"public"}]},"oa":1,"title":"Improved verification techniques for concurrent systems","month":"10","_id":"10199","user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","publication_identifier":{"issn":["2663-337X"]},"article_processing_charge":"No","file":[{"checksum":"4f412a1ee60952221b499a4b1268df35","relation":"main_file","creator":"vtoman","file_size":2915234,"file_name":"toman_th_final.pdf","access_level":"open_access","date_updated":"2021-11-08T14:12:22Z","file_id":"10225","content_type":"application/pdf","date_created":"2021-11-08T14:12:22Z"},{"checksum":"9584943f99127be2dd2963f6784c37d4","relation":"source_file","creator":"vtoman","file_size":8616056,"file_name":"toman_thesis.zip","date_updated":"2021-11-09T09:00:50Z","access_level":"closed","date_created":"2021-11-08T14:12:46Z","file_id":"10226","content_type":"application/zip"}],"acknowledged_ssus":[{"_id":"SSU"}],"project":[{"call_identifier":"H2020","name":"International IST Doctoral Program","grant_number":"665385","_id":"2564DBCA-B435-11E9-9278-68D0E5697425"},{"call_identifier":"FWF","name":"Rigorous Systems Engineering","grant_number":"S11402-N23","_id":"25F2ACDE-B435-11E9-9278-68D0E5697425"},{"grant_number":"ICT15-003","name":"Efficient Algorithms for Computer Aided Verification","_id":"25892FC0-B435-11E9-9278-68D0E5697425"},{"name":"Formal Methods for Stochastic Models: Algorithms and Applications","grant_number":"863818","_id":"0599E47C-7A3F-11EA-A408-12923DDC885E","call_identifier":"H2020"}],"date_updated":"2025-07-14T09:10:16Z","date_created":"2021-10-29T20:09:01Z","year":"2021","page":"166","author":[{"orcid":"0000-0001-9036-063X","id":"3AF3DA7C-F248-11E8-B48F-1D18A9856A87","last_name":"Toman","full_name":"Toman, Viktor","first_name":"Viktor"}],"ddc":["000"],"date_published":"2021-10-31T00:00:00Z","oa_version":"Published Version","type":"dissertation","file_date_updated":"2021-11-09T09:00:50Z","language":[{"iso":"eng"}],"publisher":"Institute of Science and Technology Austria","doi":"10.15479/at:ista:10199"},{"author":[{"last_name":"Schmid","first_name":"Laura","full_name":"Schmid, Laura","id":"38B437DE-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-6978-7329"}],"ddc":["519","576"],"date_published":"2021-11-17T00:00:00Z","oa_version":"Published Version","type":"dissertation","file_date_updated":"2022-12-20T23:30:08Z","language":[{"iso":"eng"}],"publisher":"Institute of Science and Technology Austria","doi":"10.15479/at:ista:10293","publication_identifier":{"issn":["2663-337X"]},"article_processing_charge":"No","file":[{"file_name":"submission_new.zip","access_level":"closed","date_updated":"2022-12-20T23:30:08Z","file_id":"10305","content_type":"application/zip","date_created":"2021-11-18T12:41:46Z","embargo_to":"open_access","checksum":"86a05b430756ca12ae8107b6e6f3c1e5","relation":"source_file","creator":"lschmid","file_size":29703124},{"relation":"main_file","checksum":"d940af042e94660c6b6a7b4f0b184d47","file_size":8320985,"creator":"lschmid","file_name":"thesis_new_upload.pdf","embargo":"2022-10-18","content_type":"application/pdf","file_id":"10306","date_created":"2021-11-18T12:59:15Z","access_level":"open_access","date_updated":"2022-12-20T23:30:08Z"}],"project":[{"grant_number":"279307","name":"Quantitative Graph Games: Theory and Applications","_id":"2581B60A-B435-11E9-9278-68D0E5697425","call_identifier":"FP7"},{"_id":"0599E47C-7A3F-11EA-A408-12923DDC885E","name":"Formal Methods for Stochastic Models: Algorithms and Applications","grant_number":"863818","call_identifier":"H2020"},{"call_identifier":"FWF","name":"The Wittgenstein Prize","grant_number":"Z211","_id":"25F42A32-B435-11E9-9278-68D0E5697425"},{"call_identifier":"FWF","_id":"2584A770-B435-11E9-9278-68D0E5697425","name":"Modern Graph Algorithmic Techniques in Formal Verification","grant_number":"P 23499-N23"},{"call_identifier":"FWF","_id":"25832EC2-B435-11E9-9278-68D0E5697425","name":"Rigorous Systems Engineering","grant_number":"S 11407_N23"}],"date_updated":"2025-07-14T09:10:09Z","year":"2021","date_created":"2021-11-15T17:12:57Z","page":"171","alternative_title":["ISTA Thesis"],"department":[{"_id":"GradSch"},{"_id":"KrCh"}],"publication_status":"published","related_material":{"record":[{"relation":"part_of_dissertation","status":"public","id":"9997"},{"id":"2","relation":"part_of_dissertation","status":"public"},{"id":"9402","status":"public","relation":"part_of_dissertation"}]},"title":"Evolution of cooperation via (in)direct reciprocity under imperfect information","oa":1,"month":"11","_id":"10293","user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","citation":{"short":"L. Schmid, Evolution of Cooperation via (in)Direct Reciprocity under Imperfect Information, Institute of Science and Technology Austria, 2021.","ama":"Schmid L. Evolution of cooperation via (in)direct reciprocity under imperfect information. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10293\">10.15479/at:ista:10293</a>","ista":"Schmid L. 2021. Evolution of cooperation via (in)direct reciprocity under imperfect information. Institute of Science and Technology Austria.","chicago":"Schmid, Laura. “Evolution of Cooperation via (in)Direct Reciprocity under Imperfect Information.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10293\">https://doi.org/10.15479/at:ista:10293</a>.","apa":"Schmid, L. (2021). <i>Evolution of cooperation via (in)direct reciprocity under imperfect information</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10293\">https://doi.org/10.15479/at:ista:10293</a>","ieee":"L. Schmid, “Evolution of cooperation via (in)direct reciprocity under imperfect information,” Institute of Science and Technology Austria, 2021.","mla":"Schmid, Laura. <i>Evolution of Cooperation via (in)Direct Reciprocity under Imperfect Information</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10293\">10.15479/at:ista:10293</a>."},"supervisor":[{"first_name":"Krishnendu","full_name":"Chatterjee, Krishnendu","last_name":"Chatterjee","orcid":"0000-0002-4561-241X","id":"2E5DCA20-F248-11E8-B48F-1D18A9856A87"}],"has_accepted_license":"1","day":"17","abstract":[{"text":"Indirect reciprocity in evolutionary game theory is a prominent mechanism for explaining the evolution of cooperation among unrelated individuals. In contrast to direct reciprocity, which is based on individuals meeting repeatedly, and conditionally cooperating by using their own experiences, indirect reciprocity is based on individuals’ reputations. If a player helps another, this increases the helper’s public standing, benefitting them in the future. This lets cooperation in the population emerge without individuals having to meet more than once. While the two modes of reciprocity are intertwined, they are difficult to compare. Thus, they are usually studied in isolation. Direct reciprocity can maintain cooperation with simple strategies, and is robust against noise even when players do not remember more\r\nthan their partner’s last action. Meanwhile, indirect reciprocity requires its successful strategies, or social norms, to be more complex. Exhaustive search previously identified eight such norms, called the “leading eight”, which excel at maintaining cooperation. However, as the first result of this thesis, we show that the leading eight break down once we remove the fundamental assumption that information is synchronized and public, such that everyone agrees on reputations. Once we consider a more realistic scenario of imperfect information, where reputations are private, and individuals occasionally misinterpret or miss observations, the leading eight do not promote cooperation anymore. Instead, minor initial disagreements can proliferate, fragmenting populations into subgroups. In a next step, we consider ways to mitigate this issue. We first explore whether introducing “generosity” can stabilize cooperation when players use the leading eight strategies in noisy environments. This approach of modifying strategies to include probabilistic elements for coping with errors is known to work well in direct reciprocity. However, as we show here, it fails for the more complex norms of indirect reciprocity. Imperfect information still prevents cooperation from evolving. On the other hand, we succeeded to show in this thesis that modifying the leading eight to use “quantitative assessment”, i.e. tracking reputation scores on a scale beyond good and bad, and making overall judgments of others based on a threshold, is highly successful, even when noise increases in the environment. Cooperation can flourish when reputations\r\nare more nuanced, and players have a broader understanding what it means to be “good.” Finally, we present a single theoretical framework that unites the two modes of reciprocity despite their differences. Within this framework, we identify a novel simple and successful strategy for indirect reciprocity, which can cope with noisy environments and has an analogue in direct reciprocity. We can also analyze decision making when different sources of information are available. Our results help highlight that for sustaining cooperation, already the most simple rules of reciprocity can be sufficient.","lang":"eng"}],"status":"public","ec_funded":1,"degree_awarded":"PhD"},{"abstract":[{"lang":"eng","text":"Nitrogen is an essential macronutrient determining plant growth, development and affecting agricultural productivity. Root, as a hub that perceives and integrates local and systemic signals on the plant’s external and endogenous nitrogen resources, communicates with other plant organs to consolidate their physiology and development in accordance with actual nitrogen balance. Over the last years, numerous studies demonstrated that these comprehensive developmental adaptations rely on the interaction between pathways controlling nitrogen homeostasis and hormonal networks acting globally in the plant body. However, molecular insights into how the information about the nitrogen status is translated through hormonal pathways into specific developmental output are lacking. In my work, I addressed so far poorly understood mechanisms underlying root-to-shoot communication that lead to a rapid re-adjustment of shoot growth and development after nitrate provision. Applying a combination of molecular, cell, and developmental biology approaches, genetics and grafting experiments as well as hormonal analytics, I identified and characterized an unknown molecular framework orchestrating shoot development with a root nitrate sensory system. "}],"status":"public","degree_awarded":"PhD","citation":{"ama":"Abualia R. Role of hormones in nitrate regulated growth. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10303\">10.15479/at:ista:10303</a>","short":"R. Abualia, Role of Hormones in Nitrate Regulated Growth, Institute of Science and Technology Austria, 2021.","mla":"Abualia, Rashed. <i>Role of Hormones in Nitrate Regulated Growth</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10303\">10.15479/at:ista:10303</a>.","ieee":"R. Abualia, “Role of hormones in nitrate regulated growth,” Institute of Science and Technology Austria, 2021.","apa":"Abualia, R. (2021). <i>Role of hormones in nitrate regulated growth</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10303\">https://doi.org/10.15479/at:ista:10303</a>","ista":"Abualia R. 2021. Role of hormones in nitrate regulated growth. Institute of Science and Technology Austria.","chicago":"Abualia, Rashed. “Role of Hormones in Nitrate Regulated Growth.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10303\">https://doi.org/10.15479/at:ista:10303</a>."},"supervisor":[{"last_name":"Benková","full_name":"Benková, Eva","first_name":"Eva","orcid":"0000-0002-8510-9739","id":"38F4F166-F248-11E8-B48F-1D18A9856A87"}],"has_accepted_license":"1","day":"22","related_material":{"record":[{"id":"9010","relation":"part_of_dissertation","status":"public"},{"id":"9913","status":"public","relation":"part_of_dissertation"},{"id":"47","status":"public","relation":"part_of_dissertation"}]},"title":"Role of hormones in nitrate regulated growth","oa":1,"publication_status":"published","user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","month":"11","_id":"10303","alternative_title":["ISTA Thesis"],"department":[{"_id":"GradSch"},{"_id":"EvBe"}],"file":[{"file_size":28005730,"creator":"rabualia","relation":"main_file","checksum":"dea38b98aa4da1cea03dcd0f10862818","content_type":"application/pdf","file_id":"10331","date_created":"2021-11-22T14:48:21Z","access_level":"open_access","date_updated":"2022-12-20T23:30:06Z","file_name":"AbualiaPhDthesisfinalv3.pdf","embargo":"2022-11-23"},{"creator":"rabualia","file_size":62841883,"checksum":"4cd62da5ec5ba4c32e61f0f6d9e61920","relation":"source_file","date_updated":"2022-12-20T23:30:06Z","access_level":"closed","content_type":"application/vnd.openxmlformats-officedocument.wordprocessingml.document","embargo_to":"open_access","date_created":"2021-11-22T14:48:34Z","file_id":"10332","file_name":"AbualiaPhDthesisfinalv3.docx"}],"date_updated":"2023-09-19T14:42:45Z","acknowledged_ssus":[{"_id":"LifeSc"},{"_id":"Bio"}],"page":"139","date_created":"2021-11-18T11:20:59Z","year":"2021","publication_identifier":{"issn":["2663-337X"]},"article_processing_charge":"No","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)","image":"/images/cc_by.png","short":"CC BY (4.0)"},"file_date_updated":"2022-12-20T23:30:06Z","type":"dissertation","language":[{"iso":"eng"}],"publisher":"Institute of Science and Technology Austria","doi":"10.15479/at:ista:10303","ddc":["580","581"],"author":[{"id":"4827E134-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-9357-9415","last_name":"Abualia","first_name":"Rashed","full_name":"Abualia, Rashed"}],"date_published":"2021-11-22T00:00:00Z","oa_version":"Published Version"},{"publication_identifier":{"issn":["2663-337X"]},"article_processing_charge":"No","file":[{"relation":"main_file","checksum":"b39c9e0ef18d0484d537a67551effd02","file_size":13266088,"creator":"ktomasek","file_name":"ThesisTomasekKathrin.pdf","embargo":"2022-11-18","content_type":"application/pdf","file_id":"10308","date_created":"2021-11-18T15:07:31Z","date_updated":"2022-12-20T23:30:05Z","access_level":"open_access"},{"checksum":"c0c440ee9e5ef1102a518a4f9f023e7c","relation":"source_file","creator":"ktomasek","file_size":7539509,"file_name":"ThesisTomasekKathrin.docx","access_level":"closed","date_updated":"2022-12-20T23:30:05Z","date_created":"2021-11-18T15:07:46Z","file_id":"10309","content_type":"application/vnd.openxmlformats-officedocument.wordprocessingml.document","embargo_to":"open_access"}],"acknowledged_ssus":[{"_id":"LifeSc"},{"_id":"Bio"},{"_id":"PreCl"},{"_id":"EM-Fac"}],"date_updated":"2023-09-07T13:34:38Z","date_created":"2021-11-18T15:05:06Z","year":"2021","page":"73","author":[{"last_name":"Tomasek","first_name":"Kathrin","full_name":"Tomasek, Kathrin","id":"3AEC8556-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0003-3768-877X"}],"ddc":["570"],"date_published":"2021-11-18T00:00:00Z","oa_version":"Published Version","type":"dissertation","file_date_updated":"2022-12-20T23:30:05Z","language":[{"iso":"eng"}],"publisher":"Institute of Science and Technology Austria","doi":"10.15479/at:ista:10307","citation":{"ieee":"K. Tomasek, “Pathogenic Escherichia coli hijack the host immune response,” Institute of Science and Technology Austria, 2021.","mla":"Tomasek, Kathrin. <i>Pathogenic Escherichia Coli Hijack the Host Immune Response</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10307\">10.15479/at:ista:10307</a>.","chicago":"Tomasek, Kathrin. “Pathogenic Escherichia Coli Hijack the Host Immune Response.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10307\">https://doi.org/10.15479/at:ista:10307</a>.","apa":"Tomasek, K. (2021). <i>Pathogenic Escherichia coli hijack the host immune response</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10307\">https://doi.org/10.15479/at:ista:10307</a>","ista":"Tomasek K. 2021. Pathogenic Escherichia coli hijack the host immune response. Institute of Science and Technology Austria.","ama":"Tomasek K. Pathogenic Escherichia coli hijack the host immune response. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10307\">10.15479/at:ista:10307</a>","short":"K. Tomasek, Pathogenic Escherichia Coli Hijack the Host Immune Response, Institute of Science and Technology Austria, 2021."},"supervisor":[{"last_name":"Sixt","full_name":"Sixt, Michael K","first_name":"Michael K","orcid":"0000-0002-4561-241X","id":"41E9FBEA-F248-11E8-B48F-1D18A9856A87"},{"orcid":"0000-0001-6220-2052","id":"47F8433E-F248-11E8-B48F-1D18A9856A87","full_name":"Guet, Calin C","first_name":"Calin C","last_name":"Guet"}],"has_accepted_license":"1","day":"18","abstract":[{"text":"Bacteria-host interactions represent a continuous trade-off between benefit and risk. Thus, the host immune response is faced with a non-trivial problem – accommodate beneficial commensals and remove harmful pathogens. This is especially difficult as molecular patterns, such as lipopolysaccharide or specific surface organelles such as pili, are conserved in both, commensal and pathogenic bacteria. Type 1 pili, tightly regulated by phase variation, are considered an important virulence factor of pathogenic bacteria as they facilitate invasion into host cells. While invasion represents a de facto passive mechanism for pathogens to escape the host immune response, we demonstrate a fundamental role of type 1 pili as active modulators of the innate and adaptive immune response.","lang":"eng"}],"status":"public","degree_awarded":"PhD","alternative_title":["ISTA Thesis"],"department":[{"_id":"MiSi"},{"_id":"CaGu"},{"_id":"GradSch"}],"publication_status":"published","related_material":{"record":[{"id":"10316","status":"public","relation":"part_of_dissertation"}]},"oa":1,"title":"Pathogenic Escherichia coli hijack the host immune response","month":"11","_id":"10307","user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1"},{"alternative_title":["ISTA Master's Thesis"],"author":[{"last_name":"Piankov","full_name":"Piankov, Anton","first_name":"Anton","id":"865E3C26-AA8C-11E9-A409-C4C4E5697425"}],"ddc":["530"],"oa_version":"Published Version","department":[{"_id":"GradSch"},{"_id":"CaGo"}],"date_published":"2021-12-07T00:00:00Z","language":[{"iso":"eng"}],"publication_status":"published","title":"Towards designer materials using customizable particle shape","oa":1,"type":"dissertation","file_date_updated":"2022-03-10T12:10:25Z","doi":"10.15479/at:ista:10422","_id":"10422","month":"12","user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","publisher":"Institute of Science and Technology Austria","publication_identifier":{"issn":["2791-4585"]},"supervisor":[{"last_name":"Goodrich","full_name":"Goodrich, Carl Peter","first_name":"Carl Peter","orcid":"0000-0002-1307-5074","id":"EB352CD2-F68A-11E9-89C5-A432E6697425"}],"citation":{"ieee":"A. Piankov, “Towards designer materials using customizable particle shape,” Institute of Science and Technology Austria, 2021.","mla":"Piankov, Anton. <i>Towards Designer Materials Using Customizable Particle Shape</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10422\">10.15479/at:ista:10422</a>.","chicago":"Piankov, Anton. “Towards Designer Materials Using Customizable Particle Shape.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10422\">https://doi.org/10.15479/at:ista:10422</a>.","ista":"Piankov A. 2021. Towards designer materials using customizable particle shape. Institute of Science and Technology Austria.","apa":"Piankov, A. (2021). <i>Towards designer materials using customizable particle shape</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10422\">https://doi.org/10.15479/at:ista:10422</a>","ama":"Piankov A. Towards designer materials using customizable particle shape. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10422\">10.15479/at:ista:10422</a>","short":"A. Piankov, Towards Designer Materials Using Customizable Particle Shape, Institute of Science and Technology Austria, 2021."},"day":"07","has_accepted_license":"1","article_processing_charge":"No","status":"public","file":[{"creator":"cchlebak","file_size":394018,"checksum":"114e8f4b2c002c6c352416c12de2c695","relation":"source_file","access_level":"closed","date_updated":"2022-03-10T12:10:25Z","content_type":"application/x-zip-compressed","date_created":"2021-12-07T11:13:52Z","file_id":"10424","file_name":"Thesis.zip"},{"relation":"source_file","checksum":"cd15ae991ced352a9959815f794e657c","file_size":47638,"creator":"cchlebak","file_name":"Preliminary_pages_Piankov.docx","content_type":"application/vnd.openxmlformats-officedocument.wordprocessingml.document","date_created":"2021-12-07T11:14:01Z","file_id":"10425","date_updated":"2022-03-10T12:10:25Z","access_level":"closed"},{"success":1,"access_level":"open_access","date_updated":"2021-12-07T11:20:35Z","content_type":"application/pdf","file_id":"10426","date_created":"2021-12-07T11:20:35Z","file_name":"2021_Piankov_combined.pdf","creator":"cchlebak","file_size":484965,"checksum":"e6899c798b75ba42fab9822bce309050","relation":"main_file"}],"abstract":[{"text":"Those who aim to devise new materials with desirable properties usually examine present methods first. However, they will find out that some approaches can exist only conceptually without high chances to become practically useful. It seems that a numerical technique called automatic differentiation together with increasing supply of computational accelerators will soon shift many methods of the material design from the category ”unimaginable” to the category ”expensive but possible”. Approach we suggest is not an exception. Our overall goal is to have an efficient and generalizable approach allowing to solve inverse design problems. In this thesis we scratch its surface. We consider jammed systems of identical particles. And ask ourselves how the shape of those particles (or the parameters codifying it) may affect mechanical properties of the system. An indispensable part of reaching the answer is an appropriate particle parametrization. We come up with a simple, yet generalizable and purposeful scheme for it. Using our generalizable shape parameterization, we simulate the formation of a solid composed of pentagonal-like particles and measure anisotropy in the resulting elastic response. Through automatic differentiation techniques, we directly connect the shape parameters with the elastic response. Interestingly, for our system we find that less isotropic particles lead to a more isotropic elastic response. Together with other results known about our method it seems that it can be successfully generalized for different inverse design problems.","lang":"eng"}],"year":"2021","degree_awarded":"MS","date_created":"2021-12-07T10:48:06Z","date_updated":"2023-09-07T13:34:12Z"},{"article_processing_charge":"No","publication_identifier":{"issn":["2663-337X"]},"date_updated":"2023-10-17T11:48:55Z","project":[{"call_identifier":"H2020","name":"Elastic Coordination for Scalable Machine Learning","grant_number":"805223","_id":"268A44D6-B435-11E9-9278-68D0E5697425"}],"page":"132","year":"2021","date_created":"2021-12-08T21:52:28Z","file":[{"checksum":"6bf14e9a523387328f016c0689f5e10e","relation":"main_file","creator":"gnadirad","file_size":2370859,"file_name":"Thesis_Final_09_12_2021.pdf","date_updated":"2021-12-09T17:47:49Z","success":1,"access_level":"open_access","file_id":"10436","date_created":"2021-12-09T17:47:49Z","content_type":"application/pdf"},{"file_name":"Thesis_Final_09_12_2021.zip","date_updated":"2022-03-28T12:55:12Z","access_level":"closed","date_created":"2021-12-09T17:47:49Z","file_id":"10437","content_type":"application/zip","checksum":"914d6c5ca86bd0add471971a8f4c4341","relation":"source_file","creator":"gnadirad","file_size":2596924}],"date_published":"2021-12-09T00:00:00Z","oa_version":"Published Version","ddc":["000"],"author":[{"id":"3279A00C-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0001-5634-0731","last_name":"Nadiradze","full_name":"Nadiradze, Giorgi","first_name":"Giorgi"}],"publisher":"Institute of Science and Technology Austria","doi":"10.15479/at:ista:10429","file_date_updated":"2022-03-28T12:55:12Z","type":"dissertation","language":[{"iso":"eng"}],"has_accepted_license":"1","day":"09","citation":{"ieee":"G. Nadiradze, “On achieving scalability through relaxation,” Institute of Science and Technology Austria, 2021.","mla":"Nadiradze, Giorgi. <i>On Achieving Scalability through Relaxation</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/at:ista:10429\">10.15479/at:ista:10429</a>.","ista":"Nadiradze G. 2021. On achieving scalability through relaxation. Institute of Science and Technology Austria.","apa":"Nadiradze, G. (2021). <i>On achieving scalability through relaxation</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/at:ista:10429\">https://doi.org/10.15479/at:ista:10429</a>","chicago":"Nadiradze, Giorgi. “On Achieving Scalability through Relaxation.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/at:ista:10429\">https://doi.org/10.15479/at:ista:10429</a>.","ama":"Nadiradze G. On achieving scalability through relaxation. 2021. doi:<a href=\"https://doi.org/10.15479/at:ista:10429\">10.15479/at:ista:10429</a>","short":"G. Nadiradze, On Achieving Scalability through Relaxation, Institute of Science and Technology Austria, 2021."},"supervisor":[{"last_name":"Alistarh","full_name":"Alistarh, Dan-Adrian","first_name":"Dan-Adrian","id":"4A899BFC-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0003-3650-940X"}],"ec_funded":1,"degree_awarded":"PhD","abstract":[{"text":"The scalability of concurrent data structures and distributed algorithms strongly depends on\r\nreducing the contention for shared resources and the costs of synchronization and communication. We show how such cost reductions can be attained by relaxing the strict consistency conditions required by sequential implementations. In the first part of the thesis, we consider relaxation in the context of concurrent data structures. Specifically, in data structures \r\nsuch as priority queues, imposing strong semantics renders scalability impossible, since a correct implementation of the remove operation should return only the element with highest priority. Intuitively, attempting to invoke remove operations concurrently  creates a race condition. This bottleneck  can be circumvented by relaxing semantics of the affected data structure, thus allowing removal of the elements which are no longer required to have the highest priority. We prove that the randomized implementations of relaxed data structures provide provable guarantees on the priority of the removed elements even under concurrency. Additionally, we show that in some cases the relaxed data structures can be used to scale the classical algorithms which are usually implemented with the exact ones. In the second part, we study parallel variants of the  stochastic gradient descent (SGD) algorithm, which distribute computation  among the multiple processors, thus reducing the running time. Unfortunately, in order for standard parallel SGD to succeed, each processor has to maintain a local copy of the necessary model parameter, which is identical to the local copies of other processors; the overheads from this perfect consistency in terms of communication and synchronization can negate the speedup gained by distributing the computation. We show that the consistency conditions required by SGD can be  relaxed, allowing the algorithm to be more flexible in terms of tolerating quantized communication, asynchrony, or even crash faults, while its convergence remains asymptotically the same.","lang":"eng"}],"status":"public","department":[{"_id":"GradSch"},{"_id":"DaAl"}],"alternative_title":["ISTA Thesis"],"user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","_id":"10429","month":"12","related_material":{"record":[{"id":"10432","status":"public","relation":"part_of_dissertation"},{"relation":"part_of_dissertation","status":"public","id":"6673"},{"status":"public","relation":"part_of_dissertation","id":"5965"},{"relation":"part_of_dissertation","status":"public","id":"10435"}]},"title":"On achieving scalability through relaxation","oa":1,"publication_status":"published"},{"ec_funded":1,"article_number":"e68","status":"public","abstract":[{"lang":"eng","text":"The brain efficiently performs nonlinear computations through its intricate networks of spiking neurons, but how this is done remains elusive. While nonlinear computations can be implemented successfully in spiking neural networks, this requires supervised training and the resulting connectivity can be hard to interpret. In contrast, the required connectivity for any computation in the form of a linear dynamical system can be directly derived and understood with the spike coding network (SCN) framework. These networks also have biologically realistic activity patterns and are highly robust to cell death. Here we extend the SCN framework to directly implement any polynomial dynamical system, without the need for training. This results in networks requiring a mix of synapse types (fast, slow, and multiplicative), which we term multiplicative spike coding networks (mSCNs). Using mSCNs, we demonstrate how to directly derive the required connectivity for several nonlinear dynamical systems. We also show how to carry out higher-order polynomials with coupled networks that use only pair-wise multiplicative synapses, and provide expected numbers of connections for each synapse type. Overall, our work demonstrates a novel method for implementing nonlinear computations in spiking neural networks, while keeping the attractive features of standard SCNs (robustness, realistic activity patterns, and interpretable connectivity). Finally, we discuss the biological plausibility of our approach, and how the high accuracy and robustness of the approach may be of interest for neuromorphic computing."}],"day":"15","has_accepted_license":"1","intvolume":"         1","citation":{"short":"M. Nardin, J.W. Phillips, W.F. Podlaski, S.W. Keemink, Peer Community Journal 1 (2021).","ama":"Nardin M, Phillips JW, Podlaski WF, Keemink SW. Nonlinear computations in spiking neural networks through multiplicative synapses. <i>Peer Community Journal</i>. 2021;1. doi:<a href=\"https://doi.org/10.24072/pcjournal.69\">10.24072/pcjournal.69</a>","apa":"Nardin, M., Phillips, J. W., Podlaski, W. F., &#38; Keemink, S. W. (2021). Nonlinear computations in spiking neural networks through multiplicative synapses. <i>Peer Community Journal</i>. Centre Mersenne ; Peer Community In. <a href=\"https://doi.org/10.24072/pcjournal.69\">https://doi.org/10.24072/pcjournal.69</a>","chicago":"Nardin, Michele, James W. Phillips, William F. Podlaski, and Sander W. Keemink. “Nonlinear Computations in Spiking Neural Networks through Multiplicative Synapses.” <i>Peer Community Journal</i>. Centre Mersenne ; Peer Community In, 2021. <a href=\"https://doi.org/10.24072/pcjournal.69\">https://doi.org/10.24072/pcjournal.69</a>.","ista":"Nardin M, Phillips JW, Podlaski WF, Keemink SW. 2021. Nonlinear computations in spiking neural networks through multiplicative synapses. Peer Community Journal. 1, e68.","mla":"Nardin, Michele, et al. “Nonlinear Computations in Spiking Neural Networks through Multiplicative Synapses.” <i>Peer Community Journal</i>, vol. 1, e68, Centre Mersenne ; Peer Community In, 2021, doi:<a href=\"https://doi.org/10.24072/pcjournal.69\">10.24072/pcjournal.69</a>.","ieee":"M. Nardin, J. W. Phillips, W. F. Podlaski, and S. W. Keemink, “Nonlinear computations in spiking neural networks through multiplicative synapses,” <i>Peer Community Journal</i>, vol. 1. Centre Mersenne ; Peer Community In, 2021."},"user_id":"8b945eb4-e2f2-11eb-945a-df72226e66a9","_id":"10635","month":"12","title":"Nonlinear computations in spiking neural networks through multiplicative synapses","oa":1,"publication_status":"published","article_type":"original","department":[{"_id":"GradSch"},{"_id":"JoCs"}],"acknowledgement":"A preprint version of this article has been peer-reviewed and recommended by Peer Community In Neuroscience (DOI link to the recommendation: https://doi.org/10.24072/pci.cneuro.100003).\r\nWe thank Christian Machens and Nuno Calaim for useful discussions on the project. This report\r\ncame out of a collaboration started at the CAJAL Advanced Neuroscience Training Programme in\r\nComputational Neuroscience in Lisbon, Portugal, during the 2019 summer. The authors would\r\nlike to thank the participants, TAs, lecturers, and organizers of the summer school. SWK was\r\nsupported by the Simons Collaboration on the Global Brain (543009). WFP was supported by\r\nFCT (032077). MN was supported by European Union Horizon 2020 (665385).\r\n","quality_controlled":"1","year":"2021","date_created":"2022-01-17T11:12:40Z","date_updated":"2022-01-17T13:30:01Z","project":[{"grant_number":"665385","name":"International IST Doctoral Program","_id":"2564DBCA-B435-11E9-9278-68D0E5697425","call_identifier":"H2020"}],"volume":1,"file":[{"relation":"main_file","checksum":"cd9af6b331918608f2e3d1c7940cbf4f","file_size":3311494,"creator":"mnardin","file_name":"10_24072_pcjournal_69.pdf","date_created":"2022-01-17T11:15:26Z","content_type":"application/pdf","file_id":"10636","access_level":"open_access","date_updated":"2022-01-17T11:15:26Z","success":1}],"tmp":{"legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)","image":"/images/cc_by.png","short":"CC BY (4.0)"},"article_processing_charge":"No","publication_identifier":{"eissn":["2804-3871"]},"doi":"10.24072/pcjournal.69","publisher":"Centre Mersenne ; Peer Community In","arxiv":1,"language":[{"iso":"eng"}],"file_date_updated":"2022-01-17T11:15:26Z","type":"journal_article","external_id":{"arxiv":["2009.03857"]},"oa_version":"Published Version","date_published":"2021-12-15T00:00:00Z","publication":"Peer Community Journal","ddc":["519"],"author":[{"last_name":"Nardin","full_name":"Nardin, Michele","first_name":"Michele","id":"30BD0376-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0001-8849-6570"},{"last_name":"Phillips","first_name":"James W.","full_name":"Phillips, James W."},{"last_name":"Podlaski","full_name":"Podlaski, William F.","first_name":"William F."},{"first_name":"Sander W.","full_name":"Keemink, Sander W.","last_name":"Keemink"}]},{"conference":{"location":"Virtual","end_date":"2021-02-09","name":"AAAI: Association for the Advancement of Artificial Intelligence","start_date":"2021-02-02"},"oa_version":"Published Version","external_id":{"arxiv":["2012.08185"]},"date_published":"2021-05-28T00:00:00Z","issue":"5A","publication":"Proceedings of the AAAI Conference on Artificial Intelligence","author":[{"id":"40876CD8-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-2985-7724","first_name":"Thomas A","full_name":"Henzinger, Thomas A","last_name":"Henzinger"},{"last_name":"Lechner","first_name":"Mathias","full_name":"Lechner, Mathias","id":"3DC22916-F248-11E8-B48F-1D18A9856A87"},{"id":"294AA7A6-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-4681-1699","first_name":"Dorde","full_name":"Zikelic, Dorde","last_name":"Zikelic"}],"ddc":["000"],"publisher":"AAAI Press","language":[{"iso":"eng"}],"arxiv":1,"type":"conference","file_date_updated":"2022-01-26T07:41:16Z","article_processing_charge":"No","publication_identifier":{"isbn":["978-1-57735-866-4"],"issn":["2159-5399"],"eissn":["2374-3468"]},"main_file_link":[{"open_access":"1","url":"https://ojs.aaai.org/index.php/AAAI/article/view/16496"}],"scopus_import":"1","year":"2021","date_created":"2022-01-25T15:15:02Z","quality_controlled":"1","page":"3787-3795","project":[{"call_identifier":"H2020","name":"International IST Doctoral Program","grant_number":"665385","_id":"2564DBCA-B435-11E9-9278-68D0E5697425"},{"grant_number":"Z211","name":"The Wittgenstein Prize","_id":"25F42A32-B435-11E9-9278-68D0E5697425","call_identifier":"FWF"},{"call_identifier":"H2020","_id":"0599E47C-7A3F-11EA-A408-12923DDC885E","grant_number":"863818","name":"Formal Methods for Stochastic Models: Algorithms and Applications"}],"date_updated":"2025-07-14T09:10:11Z","file":[{"file_name":"16496-Article Text-19990-1-2-20210518 (1).pdf","file_id":"10684","content_type":"application/pdf","date_created":"2022-01-26T07:41:16Z","access_level":"open_access","success":1,"date_updated":"2022-01-26T07:41:16Z","relation":"main_file","checksum":"2bc8155b2526a70fba5b7301bc89dbd1","file_size":137235,"creator":"mlechner"}],"volume":35,"department":[{"_id":"GradSch"},{"_id":"ToHe"}],"acknowledgement":"This research was supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein\r\nAward), ERC CoG 863818 (FoRM-SMArt), and the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie Grant Agreement No. 665385.\r\n","alternative_title":["Technical Tracks"],"month":"05","_id":"10665","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","publication_status":"published","oa":1,"related_material":{"record":[{"relation":"dissertation_contains","status":"public","id":"11362"}]},"title":"Scalable verification of quantized neural networks","day":"28","has_accepted_license":"1","citation":{"chicago":"Henzinger, Thomas A, Mathias Lechner, and Dorde Zikelic. “Scalable Verification of Quantized Neural Networks.” In <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, 35:3787–95. AAAI Press, 2021.","ista":"Henzinger TA, Lechner M, Zikelic D. 2021. Scalable verification of quantized neural networks. Proceedings of the AAAI Conference on Artificial Intelligence. AAAI: Association for the Advancement of Artificial Intelligence, Technical Tracks, vol. 35, 3787–3795.","apa":"Henzinger, T. A., Lechner, M., &#38; Zikelic, D. (2021). Scalable verification of quantized neural networks. In <i>Proceedings of the AAAI Conference on Artificial Intelligence</i> (Vol. 35, pp. 3787–3795). Virtual: AAAI Press.","ieee":"T. A. Henzinger, M. Lechner, and D. Zikelic, “Scalable verification of quantized neural networks,” in <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, Virtual, 2021, vol. 35, no. 5A, pp. 3787–3795.","mla":"Henzinger, Thomas A., et al. “Scalable Verification of Quantized Neural Networks.” <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, vol. 35, no. 5A, AAAI Press, 2021, pp. 3787–95.","short":"T.A. Henzinger, M. Lechner, D. Zikelic, in:, Proceedings of the AAAI Conference on Artificial Intelligence, AAAI Press, 2021, pp. 3787–3795.","ama":"Henzinger TA, Lechner M, Zikelic D. Scalable verification of quantized neural networks. In: <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>. Vol 35. AAAI Press; 2021:3787-3795."},"intvolume":"        35","ec_funded":1,"status":"public","abstract":[{"lang":"eng","text":"Formal verification of neural networks is an active topic of research, and recent advances have significantly increased the size of the networks that verification tools can handle. However, most methods are designed for verification of an idealized model of the actual network which works over real arithmetic and ignores rounding imprecisions. This idealization is in stark contrast to network quantization, which is a technique that trades numerical precision for computational efficiency and is, therefore, often applied in practice. Neglecting rounding errors of such low-bit quantized neural networks has been shown to lead to wrong conclusions about the network’s correctness. Thus, the desired approach for verifying quantized neural networks would be one that takes these rounding errors\r\ninto account. In this paper, we show that verifying the bitexact implementation of quantized neural networks with bitvector specifications is PSPACE-hard, even though verifying idealized real-valued networks and satisfiability of bit-vector specifications alone are each in NP. Furthermore, we explore several practical heuristics toward closing the complexity gap between idealized and bit-exact verification. In particular, we propose three techniques for making SMT-based verification of quantized neural networks more scalable. Our experiments demonstrate that our proposed methods allow a speedup of up to three orders of magnitude over existing approaches."}]},{"citation":{"ama":"Lechner M, Hasani R, Grosu R, Rus D, Henzinger TA. Adversarial training is not ready for robot learning. In: <i>2021 IEEE International Conference on Robotics and Automation</i>. ICRA. ; 2021:4140-4147. doi:<a href=\"https://doi.org/10.1109/ICRA48506.2021.9561036\">10.1109/ICRA48506.2021.9561036</a>","short":"M. Lechner, R. Hasani, R. Grosu, D. Rus, T.A. Henzinger, in:, 2021 IEEE International Conference on Robotics and Automation, 2021, pp. 4140–4147.","mla":"Lechner, Mathias, et al. “Adversarial Training Is Not Ready for Robot Learning.” <i>2021 IEEE International Conference on Robotics and Automation</i>, 2021, pp. 4140–47, doi:<a href=\"https://doi.org/10.1109/ICRA48506.2021.9561036\">10.1109/ICRA48506.2021.9561036</a>.","ieee":"M. Lechner, R. Hasani, R. Grosu, D. Rus, and T. A. Henzinger, “Adversarial training is not ready for robot learning,” in <i>2021 IEEE International Conference on Robotics and Automation</i>, Xi’an, China, 2021, pp. 4140–4147.","ista":"Lechner M, Hasani R, Grosu R, Rus D, Henzinger TA. 2021. Adversarial training is not ready for robot learning. 2021 IEEE International Conference on Robotics and Automation. ICRA: International Conference on Robotics and AutomationICRA, 4140–4147.","apa":"Lechner, M., Hasani, R., Grosu, R., Rus, D., &#38; Henzinger, T. A. (2021). Adversarial training is not ready for robot learning. In <i>2021 IEEE International Conference on Robotics and Automation</i> (pp. 4140–4147). Xi’an, China. <a href=\"https://doi.org/10.1109/ICRA48506.2021.9561036\">https://doi.org/10.1109/ICRA48506.2021.9561036</a>","chicago":"Lechner, Mathias, Ramin Hasani, Radu Grosu, Daniela Rus, and Thomas A Henzinger. “Adversarial Training Is Not Ready for Robot Learning.” In <i>2021 IEEE International Conference on Robotics and Automation</i>, 4140–47. ICRA, 2021. <a href=\"https://doi.org/10.1109/ICRA48506.2021.9561036\">https://doi.org/10.1109/ICRA48506.2021.9561036</a>."},"has_accepted_license":"1","abstract":[{"text":"Adversarial training is an effective method to train deep learning models that are resilient to norm-bounded perturbations, with the cost of nominal performance drop. While adversarial training appears to enhance the robustness and safety of a deep model deployed in open-world decision-critical applications, counterintuitively, it induces undesired behaviors in robot learning settings. In this paper, we show theoretically and experimentally that neural controllers obtained via adversarial training are subjected to three types of defects, namely transient, systematic, and conditional errors. We first generalize adversarial training to a safety-domain optimization scheme allowing for more generic specifications. We then prove that such a learning process tends to cause certain error profiles. We support our theoretical results by a thorough experimental safety analysis in a robot-learning task. Our results suggest that adversarial training is not yet ready for robot learning.","lang":"eng"}],"status":"public","acknowledgement":"M.L. and T.A.H. are supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award). R.H. and D.R. are supported by Boeing and R.G. by Horizon-2020 ECSEL Project grant no. 783163 (iDev40).","department":[{"_id":"GradSch"},{"_id":"ToHe"}],"oa":1,"related_material":{"record":[{"status":"public","relation":"dissertation_contains","id":"11362"}]},"title":"Adversarial training is not ready for robot learning","publication_status":"published","user_id":"4359f0d1-fa6c-11eb-b949-802e58b17ae8","_id":"10666","main_file_link":[{"open_access":"1","url":"https://arxiv.org/abs/2103.08187"}],"publication_identifier":{"isbn":["978-1-7281-9078-5"],"eisbn":["978-1-7281-9077-8"],"eissn":["2577-087X"],"issn":["1050-4729"]},"article_processing_charge":"No","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by-nc-nd/3.0/legalcode","image":"/images/cc_by_nc_nd.png","name":"Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported (CC BY-NC-ND 3.0)","short":"CC BY-NC-ND (3.0)"},"isi":1,"date_updated":"2023-08-17T06:58:38Z","project":[{"grant_number":"Z211","name":"The Wittgenstein Prize","_id":"25F42A32-B435-11E9-9278-68D0E5697425","call_identifier":"FWF"}],"page":"4140-4147","quality_controlled":"1","date_created":"2022-01-25T15:44:54Z","year":"2021","series_title":"ICRA","ddc":["000"],"author":[{"id":"3DC22916-F248-11E8-B48F-1D18A9856A87","last_name":"Lechner","full_name":"Lechner, Mathias","first_name":"Mathias"},{"last_name":"Hasani","full_name":"Hasani, Ramin","first_name":"Ramin"},{"last_name":"Grosu","first_name":"Radu","full_name":"Grosu, Radu"},{"full_name":"Rus, Daniela","first_name":"Daniela","last_name":"Rus"},{"id":"40876CD8-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-2985-7724","last_name":"Henzinger","first_name":"Thomas A","full_name":"Henzinger, Thomas A"}],"publication":"2021 IEEE International Conference on Robotics and Automation","date_published":"2021-01-01T00:00:00Z","license":"https://creativecommons.org/licenses/by-nc-nd/3.0/","external_id":{"arxiv":["2103.08187"],"isi":["000765738803040"]},"oa_version":"None","conference":{"name":"ICRA: International Conference on Robotics and Automation","start_date":"2021-05-30","location":"Xi'an, China","end_date":"2021-06-05"},"type":"conference","language":[{"iso":"eng"}],"arxiv":1,"doi":"10.1109/ICRA48506.2021.9561036"},{"doi":"10.48550/arXiv.2111.03165","type":"conference","file_date_updated":"2022-01-26T07:39:59Z","arxiv":1,"language":[{"iso":"eng"}],"date_published":"2021-12-01T00:00:00Z","external_id":{"arxiv":["2111.03165"]},"conference":{"end_date":"2021-12-10","location":"Virtual","start_date":"2021-12-06","name":"NeurIPS: Neural Information Processing Systems"},"oa_version":"Published Version","author":[{"last_name":"Lechner","full_name":"Lechner, Mathias","first_name":"Mathias","id":"3DC22916-F248-11E8-B48F-1D18A9856A87"},{"last_name":"Žikelić","first_name":"Ðorđe","full_name":"Žikelić, Ðorđe"},{"last_name":"Chatterjee","full_name":"Chatterjee, Krishnendu","first_name":"Krishnendu","orcid":"0000-0002-4561-241X","id":"2E5DCA20-F248-11E8-B48F-1D18A9856A87"},{"id":"40876CD8-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0002-2985-7724","last_name":"Henzinger","full_name":"Henzinger, Thomas A","first_name":"Thomas A"}],"ddc":["000"],"publication":"35th Conference on Neural Information Processing Systems","project":[{"call_identifier":"H2020","_id":"2564DBCA-B435-11E9-9278-68D0E5697425","name":"International IST Doctoral Program","grant_number":"665385"},{"call_identifier":"H2020","_id":"0599E47C-7A3F-11EA-A408-12923DDC885E","grant_number":"863818","name":"Formal Methods for Stochastic Models: Algorithms and Applications"},{"call_identifier":"FWF","grant_number":"Z211","name":"The Wittgenstein Prize","_id":"25F42A32-B435-11E9-9278-68D0E5697425"}],"date_updated":"2025-07-14T09:10:12Z","year":"2021","date_created":"2022-01-25T15:45:58Z","quality_controlled":"1","file":[{"creator":"mlechner","file_size":452492,"checksum":"0fc0f852525c10dda9cc9ffea07fb4e4","relation":"main_file","success":1,"date_updated":"2022-01-26T07:39:59Z","access_level":"open_access","date_created":"2022-01-26T07:39:59Z","file_id":"10682","content_type":"application/pdf","file_name":"infinite_time_horizon_safety_o.pdf"}],"tmp":{"legal_code_url":"https://creativecommons.org/licenses/by-nc-nd/3.0/legalcode","image":"/images/cc_by_nc_nd.png","name":"Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported (CC BY-NC-ND 3.0)","short":"CC BY-NC-ND (3.0)"},"article_processing_charge":"No","main_file_link":[{"open_access":"1","url":"https://proceedings.neurips.cc/paper/2021/hash/544defa9fddff50c53b71c43e0da72be-Abstract.html"}],"month":"12","_id":"10667","user_id":"2EBD1598-F248-11E8-B48F-1D18A9856A87","publication_status":"published","related_material":{"record":[{"relation":"dissertation_contains","status":"public","id":"11362"}]},"oa":1,"title":"Infinite time horizon safety of Bayesian neural networks","department":[{"_id":"GradSch"},{"_id":"ToHe"},{"_id":"KrCh"}],"acknowledgement":"This research was supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award), ERC CoG 863818 (FoRM-SMArt), and the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie Grant Agreement No. 665385.","alternative_title":[" Advances in Neural Information Processing Systems"],"ec_funded":1,"abstract":[{"text":"Bayesian neural networks (BNNs) place distributions over the weights of a neural network to model uncertainty in the data and the network's prediction. We consider the problem of verifying safety when running a Bayesian neural network policy in a feedback loop with infinite time horizon systems. Compared to the existing sampling-based approaches, which are inapplicable to the infinite time horizon setting, we train a separate deterministic neural network that serves as an infinite time horizon safety certificate. In particular, we show that the certificate network guarantees the safety of the system over a subset of the BNN weight posterior's support. Our method first computes a safe weight set and then alters the BNN's weight posterior to reject samples outside this set. Moreover, we show how to extend our approach to a safe-exploration reinforcement learning setting, in order to avoid unsafe trajectories during the training of the policy. We evaluate our approach on a series of reinforcement learning benchmarks, including non-Lyapunovian safety specifications.","lang":"eng"}],"status":"public","has_accepted_license":"1","day":"01","citation":{"short":"M. Lechner, Ð. Žikelić, K. Chatterjee, T.A. Henzinger, in:, 35th Conference on Neural Information Processing Systems, 2021.","ama":"Lechner M, Žikelić Ð, Chatterjee K, Henzinger TA. Infinite time horizon safety of Bayesian neural networks. In: <i>35th Conference on Neural Information Processing Systems</i>. ; 2021. doi:<a href=\"https://doi.org/10.48550/arXiv.2111.03165\">10.48550/arXiv.2111.03165</a>","ista":"Lechner M, Žikelić Ð, Chatterjee K, Henzinger TA. 2021. Infinite time horizon safety of Bayesian neural networks. 35th Conference on Neural Information Processing Systems. NeurIPS: Neural Information Processing Systems,  Advances in Neural Information Processing Systems, .","chicago":"Lechner, Mathias, Ðorđe Žikelić, Krishnendu Chatterjee, and Thomas A Henzinger. “Infinite Time Horizon Safety of Bayesian Neural Networks.” In <i>35th Conference on Neural Information Processing Systems</i>, 2021. <a href=\"https://doi.org/10.48550/arXiv.2111.03165\">https://doi.org/10.48550/arXiv.2111.03165</a>.","apa":"Lechner, M., Žikelić, Ð., Chatterjee, K., &#38; Henzinger, T. A. (2021). Infinite time horizon safety of Bayesian neural networks. In <i>35th Conference on Neural Information Processing Systems</i>. Virtual. <a href=\"https://doi.org/10.48550/arXiv.2111.03165\">https://doi.org/10.48550/arXiv.2111.03165</a>","ieee":"M. Lechner, Ð. Žikelić, K. Chatterjee, and T. A. Henzinger, “Infinite time horizon safety of Bayesian neural networks,” in <i>35th Conference on Neural Information Processing Systems</i>, Virtual, 2021.","mla":"Lechner, Mathias, et al. “Infinite Time Horizon Safety of Bayesian Neural Networks.” <i>35th Conference on Neural Information Processing Systems</i>, 2021, doi:<a href=\"https://doi.org/10.48550/arXiv.2111.03165\">10.48550/arXiv.2111.03165</a>."}},{"year":"2021","date_created":"2022-01-25T15:46:33Z","quality_controlled":"1","page":"478-489","project":[{"call_identifier":"FWF","name":"The Wittgenstein Prize","grant_number":"Z211","_id":"25F42A32-B435-11E9-9278-68D0E5697425"}],"date_updated":"2022-05-04T15:02:27Z","file":[{"file_size":4246561,"creator":"mlechner","relation":"main_file","checksum":"d30eae62561bb517d9f978437d7677db","date_created":"2022-01-26T07:38:32Z","content_type":"application/pdf","file_id":"10681","success":1,"access_level":"open_access","date_updated":"2022-01-26T07:38:32Z","file_name":"babaiee21a.pdf"}],"volume":139,"article_processing_charge":"No","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by-nc-nd/3.0/legalcode","image":"/images/cc_by_nc_nd.png","name":"Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported (CC BY-NC-ND 3.0)","short":"CC BY-NC-ND (3.0)"},"publication_identifier":{"issn":["2640-3498"]},"main_file_link":[{"url":"https://proceedings.mlr.press/v139/babaiee21a","open_access":"1"}],"publisher":"ML Research Press","language":[{"iso":"eng"}],"type":"conference","file_date_updated":"2022-01-26T07:38:32Z","conference":{"end_date":"2021-07-24","location":"Virtual","start_date":"2021-07-18","name":"ML: Machine Learning"},"oa_version":"Published Version","date_published":"2021-07-01T00:00:00Z","publication":"Proceedings of the 38th International Conference on Machine Learning","author":[{"full_name":"Babaiee, Zahra","first_name":"Zahra","last_name":"Babaiee"},{"first_name":"Ramin","full_name":"Hasani, Ramin","last_name":"Hasani"},{"full_name":"Lechner, Mathias","first_name":"Mathias","last_name":"Lechner","id":"3DC22916-F248-11E8-B48F-1D18A9856A87"},{"first_name":"Daniela","full_name":"Rus, Daniela","last_name":"Rus"},{"last_name":"Grosu","full_name":"Grosu, Radu","first_name":"Radu"}],"ddc":["000"],"status":"public","abstract":[{"text":"Robustness to variations in lighting conditions is a key objective for any deep vision system. To this end, our paper extends the receptive field of convolutional neural networks with two residual components, ubiquitous in the visual processing system of vertebrates: On-center and off-center pathways, with an excitatory center and inhibitory surround; OOCS for short. The On-center pathway is excited by the presence of a light stimulus in its center, but not in its surround, whereas the Off-center pathway is excited by the absence of a light stimulus in its center, but not in its surround. We design OOCS pathways via a difference of Gaussians, with their variance computed analytically from the size of the receptive fields. OOCS pathways complement each other in their response to light stimuli, ensuring this way a strong edge-detection capability, and as a result an accurate and robust inference under challenging lighting conditions. We provide extensive empirical evidence showing that networks supplied with OOCS pathways gain accuracy and illumination-robustness from the novel edge representation, compared to other baselines.","lang":"eng"}],"day":"01","has_accepted_license":"1","intvolume":"       139","citation":{"ama":"Babaiee Z, Hasani R, Lechner M, Rus D, Grosu R. On-off center-surround receptive fields for accurate and robust image classification. In: <i>Proceedings of the 38th International Conference on Machine Learning</i>. Vol 139. ML Research Press; 2021:478-489.","short":"Z. Babaiee, R. Hasani, M. Lechner, D. Rus, R. Grosu, in:, Proceedings of the 38th International Conference on Machine Learning, ML Research Press, 2021, pp. 478–489.","ieee":"Z. Babaiee, R. Hasani, M. Lechner, D. Rus, and R. Grosu, “On-off center-surround receptive fields for accurate and robust image classification,” in <i>Proceedings of the 38th International Conference on Machine Learning</i>, Virtual, 2021, vol. 139, pp. 478–489.","mla":"Babaiee, Zahra, et al. “On-off Center-Surround Receptive Fields for Accurate and Robust Image Classification.” <i>Proceedings of the 38th International Conference on Machine Learning</i>, vol. 139, ML Research Press, 2021, pp. 478–89.","chicago":"Babaiee, Zahra, Ramin Hasani, Mathias Lechner, Daniela Rus, and Radu Grosu. “On-off Center-Surround Receptive Fields for Accurate and Robust Image Classification.” In <i>Proceedings of the 38th International Conference on Machine Learning</i>, 139:478–89. ML Research Press, 2021.","ista":"Babaiee Z, Hasani R, Lechner M, Rus D, Grosu R. 2021. On-off center-surround receptive fields for accurate and robust image classification. Proceedings of the 38th International Conference on Machine Learning. ML: Machine Learning, PMLR, vol. 139, 478–489.","apa":"Babaiee, Z., Hasani, R., Lechner, M., Rus, D., &#38; Grosu, R. (2021). On-off center-surround receptive fields for accurate and robust image classification. In <i>Proceedings of the 38th International Conference on Machine Learning</i> (Vol. 139, pp. 478–489). Virtual: ML Research Press."},"_id":"10668","month":"07","user_id":"2EBD1598-F248-11E8-B48F-1D18A9856A87","publication_status":"published","title":"On-off center-surround receptive fields for accurate and robust image classification","oa":1,"acknowledgement":"Z.B. is supported by the Doctoral College Resilient Embedded Systems, which is run jointly by the TU Wien’s Faculty of Informatics and the UAS Technikum Wien. R.G. is partially supported by the Horizon 2020 Era-Permed project Persorad, and ECSEL Project grant no. 783163 (iDev40). R.H and D.R were partially supported by Boeing and MIT. M.L. is supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award).","department":[{"_id":"GradSch"},{"_id":"ToHe"}],"alternative_title":["PMLR"]},{"publisher":"AAAI Press","file_date_updated":"2022-01-26T07:38:08Z","type":"conference","arxiv":1,"language":[{"iso":"eng"}],"date_published":"2021-05-28T00:00:00Z","conference":{"name":"AAAI: Association for the Advancement of Artificial Intelligence","start_date":"2021-02-02","location":"Virtual","end_date":"2021-02-09"},"oa_version":"Published Version","external_id":{"arxiv":["2012.08863"]},"ddc":["000"],"author":[{"last_name":"Grunbacher","first_name":"Sophie","full_name":"Grunbacher, Sophie"},{"last_name":"Hasani","full_name":"Hasani, Ramin","first_name":"Ramin"},{"last_name":"Lechner","first_name":"Mathias","full_name":"Lechner, Mathias","id":"3DC22916-F248-11E8-B48F-1D18A9856A87"},{"first_name":"Jacek","full_name":"Cyranka, Jacek","last_name":"Cyranka"},{"first_name":"Scott A","full_name":"Smolka, Scott A","last_name":"Smolka"},{"last_name":"Grosu","first_name":"Radu","full_name":"Grosu, Radu"}],"publication":"Proceedings of the AAAI Conference on Artificial Intelligence","issue":"13","date_updated":"2022-05-24T06:33:14Z","project":[{"call_identifier":"FWF","_id":"25F42A32-B435-11E9-9278-68D0E5697425","name":"The Wittgenstein Prize","grant_number":"Z211"}],"page":"11525-11535","quality_controlled":"1","year":"2021","date_created":"2022-01-25T15:47:20Z","volume":35,"file":[{"access_level":"open_access","date_updated":"2022-01-26T07:38:08Z","success":1,"file_id":"10680","content_type":"application/pdf","date_created":"2022-01-26T07:38:08Z","file_name":"17372-Article Text-20866-1-2-20210518.pdf","creator":"mlechner","file_size":286906,"checksum":"468d07041e282a1d46ffdae92f709630","relation":"main_file"}],"article_processing_charge":"No","main_file_link":[{"url":"https://ojs.aaai.org/index.php/AAAI/article/view/17372","open_access":"1"}],"publication_identifier":{"issn":["2159-5399"],"eissn":["2374-3468"],"isbn":["978-1-57735-866-4"]},"user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","month":"05","_id":"10669","title":"On the verification of neural ODEs with stochastic guarantees","oa":1,"publication_status":"published","department":[{"_id":"GradSch"},{"_id":"ToHe"}],"acknowledgement":"The authors would like to thank the reviewers for their insightful comments. RH and RG were partially supported by\r\nHorizon-2020 ECSEL Project grant No. 783163 (iDev40). RH was partially supported by Boeing. ML was supported\r\nin part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award). SG was funded by FWF\r\nproject W1255-N23. JC was partially supported by NAWA Polish Returns grant PPN/PPO/2018/1/00029. SS was supported by NSF awards DCL-2040599, CCF-1918225, and CPS-1446832.\r\n","alternative_title":["Technical Tracks"],"abstract":[{"lang":"eng","text":"We show that Neural ODEs, an emerging class of timecontinuous neural networks, can be verified by solving a set of global-optimization problems. For this purpose, we introduce Stochastic Lagrangian Reachability (SLR), an\r\nabstraction-based technique for constructing a tight Reachtube (an over-approximation of the set of reachable states\r\nover a given time-horizon), and provide stochastic guarantees in the form of confidence intervals for the Reachtube bounds. SLR inherently avoids the infamous wrapping effect (accumulation of over-approximation errors) by performing local optimization steps to expand safe regions instead of repeatedly forward-propagating them as is done by deterministic reachability methods. To enable fast local optimizations, we introduce a novel forward-mode adjoint sensitivity method to compute gradients without the need for backpropagation. Finally, we establish asymptotic and non-asymptotic convergence rates for SLR."}],"status":"public","has_accepted_license":"1","day":"28","intvolume":"        35","citation":{"ama":"Grunbacher S, Hasani R, Lechner M, Cyranka J, Smolka SA, Grosu R. On the verification of neural ODEs with stochastic guarantees. In: <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>. Vol 35. AAAI Press; 2021:11525-11535.","short":"S. Grunbacher, R. Hasani, M. Lechner, J. Cyranka, S.A. Smolka, R. Grosu, in:, Proceedings of the AAAI Conference on Artificial Intelligence, AAAI Press, 2021, pp. 11525–11535.","ieee":"S. Grunbacher, R. Hasani, M. Lechner, J. Cyranka, S. A. Smolka, and R. Grosu, “On the verification of neural ODEs with stochastic guarantees,” in <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, Virtual, 2021, vol. 35, no. 13, pp. 11525–11535.","mla":"Grunbacher, Sophie, et al. “On the Verification of Neural ODEs with Stochastic Guarantees.” <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, vol. 35, no. 13, AAAI Press, 2021, pp. 11525–35.","apa":"Grunbacher, S., Hasani, R., Lechner, M., Cyranka, J., Smolka, S. A., &#38; Grosu, R. (2021). On the verification of neural ODEs with stochastic guarantees. In <i>Proceedings of the AAAI Conference on Artificial Intelligence</i> (Vol. 35, pp. 11525–11535). Virtual: AAAI Press.","chicago":"Grunbacher, Sophie, Ramin Hasani, Mathias Lechner, Jacek Cyranka, Scott A Smolka, and Radu Grosu. “On the Verification of Neural ODEs with Stochastic Guarantees.” In <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, 35:11525–35. AAAI Press, 2021.","ista":"Grunbacher S, Hasani R, Lechner M, Cyranka J, Smolka SA, Grosu R. 2021. On the verification of neural ODEs with stochastic guarantees. Proceedings of the AAAI Conference on Artificial Intelligence. AAAI: Association for the Advancement of Artificial Intelligence, Technical Tracks, vol. 35, 11525–11535."}},{"project":[{"call_identifier":"FWF","grant_number":"Z211","name":"The Wittgenstein Prize","_id":"25F42A32-B435-11E9-9278-68D0E5697425"}],"date_updated":"2022-01-26T14:33:31Z","date_created":"2022-01-25T15:47:50Z","year":"2021","quality_controlled":"1","file":[{"file_name":"NeurIPS-2021-causal-navigation-by-continuous-time-neural-networks-Paper.pdf","access_level":"open_access","date_updated":"2022-01-26T07:37:24Z","success":1,"date_created":"2022-01-26T07:37:24Z","file_id":"10679","content_type":"application/pdf","checksum":"be81f0ade174a8c9b2d4fe09590b2021","relation":"main_file","creator":"mlechner","file_size":6841228}],"article_processing_charge":"No","tmp":{"legal_code_url":"https://creativecommons.org/licenses/by-nc-nd/3.0/legalcode","image":"/images/cc_by_nc_nd.png","name":"Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported (CC BY-NC-ND 3.0)","short":"CC BY-NC-ND (3.0)"},"main_file_link":[{"url":"https://proceedings.neurips.cc/paper/2021/hash/67ba02d73c54f0b83c05507b7fb7267f-Abstract.html","open_access":"1"}],"type":"conference","file_date_updated":"2022-01-26T07:37:24Z","arxiv":1,"language":[{"iso":"eng"}],"date_published":"2021-12-01T00:00:00Z","external_id":{"arxiv":["2106.08314"]},"oa_version":"Published Version","conference":{"end_date":"2021-12-10","location":"Virtual","start_date":"2021-12-06","name":"NeurIPS: Neural Information Processing Systems"},"author":[{"full_name":"Vorbach, Charles J","first_name":"Charles J","last_name":"Vorbach"},{"last_name":"Hasani","full_name":"Hasani, Ramin","first_name":"Ramin"},{"last_name":"Amini","full_name":"Amini, Alexander","first_name":"Alexander"},{"id":"3DC22916-F248-11E8-B48F-1D18A9856A87","last_name":"Lechner","full_name":"Lechner, Mathias","first_name":"Mathias"},{"last_name":"Rus","first_name":"Daniela","full_name":"Rus, Daniela"}],"ddc":["000"],"publication":"35th Conference on Neural Information Processing Systems","abstract":[{"lang":"eng","text":"Imitation learning enables high-fidelity, vision-based learning of policies within rich, photorealistic environments. However, such techniques often rely on traditional discrete-time neural models and face difficulties in generalizing to domain shifts by failing to account for the causal relationships between the agent and the environment. In this paper, we propose a theoretical and experimental framework for learning causal representations using continuous-time neural networks, specifically over their discrete-time counterparts. We evaluate our method in the context of visual-control learning of drones over a series of complex tasks, ranging from short- and long-term navigation, to chasing static and dynamic objects through photorealistic environments. Our results demonstrate that causal continuous-time\r\ndeep models can perform robust navigation tasks, where advanced recurrent models fail. These models learn complex causal control representations directly from raw visual inputs and scale to solve a variety of tasks using imitation learning."}],"status":"public","has_accepted_license":"1","day":"01","citation":{"apa":"Vorbach, C. J., Hasani, R., Amini, A., Lechner, M., &#38; Rus, D. (2021). Causal navigation by continuous-time neural networks. In <i>35th Conference on Neural Information Processing Systems</i>. Virtual.","chicago":"Vorbach, Charles J, Ramin Hasani, Alexander Amini, Mathias Lechner, and Daniela Rus. “Causal Navigation by Continuous-Time Neural Networks.” In <i>35th Conference on Neural Information Processing Systems</i>, 2021.","ista":"Vorbach CJ, Hasani R, Amini A, Lechner M, Rus D. 2021. Causal navigation by continuous-time neural networks. 35th Conference on Neural Information Processing Systems. NeurIPS: Neural Information Processing Systems,  Advances in Neural Information Processing Systems, .","ieee":"C. J. Vorbach, R. Hasani, A. Amini, M. Lechner, and D. Rus, “Causal navigation by continuous-time neural networks,” in <i>35th Conference on Neural Information Processing Systems</i>, Virtual, 2021.","mla":"Vorbach, Charles J., et al. “Causal Navigation by Continuous-Time Neural Networks.” <i>35th Conference on Neural Information Processing Systems</i>, 2021.","short":"C.J. Vorbach, R. Hasani, A. Amini, M. Lechner, D. Rus, in:, 35th Conference on Neural Information Processing Systems, 2021.","ama":"Vorbach CJ, Hasani R, Amini A, Lechner M, Rus D. Causal navigation by continuous-time neural networks. In: <i>35th Conference on Neural Information Processing Systems</i>. ; 2021."},"month":"12","_id":"10670","user_id":"8b945eb4-e2f2-11eb-945a-df72226e66a9","publication_status":"published","oa":1,"title":"Causal navigation by continuous-time neural networks","department":[{"_id":"GradSch"},{"_id":"ToHe"}],"acknowledgement":"C.V., R.H. A.A. and D.R. are partially supported by Boeing and MIT. A.A. is supported by the National Science Foundation (NSF) Graduate Research Fellowship Program. M.L. is supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award). Research was sponsored by the United States Air Force Research Laboratory and the United States Air Force Artificial Intelligence Accelerator and was accomplished under Cooperative Agreement Number FA8750-19-2-1000. The views and conclusions contained in this document are those of the authors\r\nand should not be interpreted as representing the official policies, either expressed or implied, of the United States Air Force or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein.\r\n","alternative_title":[" Advances in Neural Information Processing Systems"]},{"acknowledgement":"R.H. and D.R. are partially supported by Boeing. R.H. and R.G. were partially supported by the Horizon-2020 ECSEL\r\nProject grant No. 783163 (iDev40). M.L. was supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award). A.A. is supported by the National Science Foundation (NSF) Graduate Research Fellowship Program. This research work is partially drawn from the PhD dissertation of R.H.","department":[{"_id":"GradSch"},{"_id":"ToHe"}],"alternative_title":["Technical Tracks"],"_id":"10671","month":"05","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","publication_status":"published","oa":1,"title":"Liquid time-constant networks","has_accepted_license":"1","day":"28","intvolume":"        35","citation":{"short":"R. Hasani, M. Lechner, A. Amini, D. Rus, R. Grosu, in:, Proceedings of the AAAI Conference on Artificial Intelligence, AAAI Press, 2021, pp. 7657–7666.","ama":"Hasani R, Lechner M, Amini A, Rus D, Grosu R. Liquid time-constant networks. In: <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>. Vol 35. AAAI Press; 2021:7657-7666.","apa":"Hasani, R., Lechner, M., Amini, A., Rus, D., &#38; Grosu, R. (2021). Liquid time-constant networks. In <i>Proceedings of the AAAI Conference on Artificial Intelligence</i> (Vol. 35, pp. 7657–7666). Virtual: AAAI Press.","ista":"Hasani R, Lechner M, Amini A, Rus D, Grosu R. 2021. Liquid time-constant networks. Proceedings of the AAAI Conference on Artificial Intelligence. AAAI: Association for the Advancement of Artificial Intelligence, Technical Tracks, vol. 35, 7657–7666.","chicago":"Hasani, Ramin, Mathias Lechner, Alexander Amini, Daniela Rus, and Radu Grosu. “Liquid Time-Constant Networks.” In <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, 35:7657–66. AAAI Press, 2021.","mla":"Hasani, Ramin, et al. “Liquid Time-Constant Networks.” <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, vol. 35, no. 9, AAAI Press, 2021, pp. 7657–66.","ieee":"R. Hasani, M. Lechner, A. Amini, D. Rus, and R. Grosu, “Liquid time-constant networks,” in <i>Proceedings of the AAAI Conference on Artificial Intelligence</i>, Virtual, 2021, vol. 35, no. 9, pp. 7657–7666."},"abstract":[{"lang":"eng","text":"We introduce a new class of time-continuous recurrent neural network models. Instead of declaring a learning system’s dynamics by implicit nonlinearities, we construct networks of linear first-order dynamical systems modulated via nonlinear interlinked gates. The resulting models represent dynamical systems with varying (i.e., liquid) time-constants coupled to their hidden state, with outputs being computed by numerical differential equation solvers. These neural networks exhibit stable and bounded behavior, yield superior expressivity within the family of neural ordinary differential equations, and give rise to improved performance on time-series prediction tasks. To demonstrate these properties, we first take a theoretical approach to find bounds over their dynamics, and compute their expressive power by the trajectory length measure in a latent trajectory space. We then conduct a series of time-series prediction experiments to manifest the approximation capability of Liquid Time-Constant Networks (LTCs) compared to classical and modern RNNs."}],"status":"public","date_published":"2021-05-28T00:00:00Z","external_id":{"arxiv":["2006.04439"]},"conference":{"location":"Virtual","end_date":"2021-02-09","name":"AAAI: Association for the Advancement of Artificial Intelligence","start_date":"2021-02-02"},"oa_version":"Published Version","author":[{"last_name":"Hasani","full_name":"Hasani, Ramin","first_name":"Ramin"},{"id":"3DC22916-F248-11E8-B48F-1D18A9856A87","first_name":"Mathias","full_name":"Lechner, Mathias","last_name":"Lechner"},{"full_name":"Amini, Alexander","first_name":"Alexander","last_name":"Amini"},{"first_name":"Daniela","full_name":"Rus, Daniela","last_name":"Rus"},{"last_name":"Grosu","full_name":"Grosu, Radu","first_name":"Radu"}],"ddc":["000"],"issue":"9","publication":"Proceedings of the AAAI Conference on Artificial Intelligence","publisher":"AAAI Press","type":"conference","file_date_updated":"2022-01-26T07:36:03Z","arxiv":1,"language":[{"iso":"eng"}],"article_processing_charge":"No","main_file_link":[{"url":"https://ojs.aaai.org/index.php/AAAI/article/view/16936","open_access":"1"}],"publication_identifier":{"eissn":["2374-3468"],"issn":["2159-5399"],"isbn":["978-1-57735-866-4"]},"project":[{"_id":"25F42A32-B435-11E9-9278-68D0E5697425","grant_number":"Z211","name":"The Wittgenstein Prize","call_identifier":"FWF"}],"date_updated":"2022-05-24T06:36:54Z","year":"2021","date_created":"2022-01-25T15:48:36Z","page":"7657-7666","quality_controlled":"1","file":[{"file_size":4302669,"creator":"mlechner","relation":"main_file","checksum":"0f06995fba06dbcfa7ed965fc66027ff","content_type":"application/pdf","date_created":"2022-01-26T07:36:03Z","file_id":"10678","access_level":"open_access","date_updated":"2022-01-26T07:36:03Z","success":1,"file_name":"16936-Article Text-20430-1-2-20210518 (1).pdf"}],"volume":35},{"acknowledgement":"This research was supported in part by the Austrian Science Fund (FWF) under grant Z211-N23 (Wittgenstein Award), ERC CoG 863818 (FoRM-SMArt), and by the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie Grant Agreement No. 665385.","department":[{"_id":"GradSch"},{"_id":"KrCh"}],"_id":"10694","month":"01","user_id":"8b945eb4-e2f2-11eb-945a-df72226e66a9","publication_status":"published","title":"Infinite-duration all-pay bidding games","oa":1,"day":"01","citation":{"ieee":"G. Avni, I. R. Jecker, and D. Zikelic, “Infinite-duration all-pay bidding games,” in <i>Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms</i>, Virtual, 2021, pp. 617–636.","mla":"Avni, Guy, et al. “Infinite-Duration All-Pay Bidding Games.” <i>Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms</i>, edited by Dániel Marx, Society for Industrial and Applied Mathematics, 2021, pp. 617–36, doi:<a href=\"https://doi.org/10.1137/1.9781611976465.38\">10.1137/1.9781611976465.38</a>.","ista":"Avni G, Jecker IR, Zikelic D. 2021. Infinite-duration all-pay bidding games. Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms. SODA: Symposium on Discrete Algorithms, 617–636.","chicago":"Avni, Guy, Ismael R Jecker, and Dorde Zikelic. “Infinite-Duration All-Pay Bidding Games.” In <i>Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms</i>, edited by Dániel Marx, 617–36. Society for Industrial and Applied Mathematics, 2021. <a href=\"https://doi.org/10.1137/1.9781611976465.38\">https://doi.org/10.1137/1.9781611976465.38</a>.","apa":"Avni, G., Jecker, I. R., &#38; Zikelic, D. (2021). Infinite-duration all-pay bidding games. In D. Marx (Ed.), <i>Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms</i> (pp. 617–636). Virtual: Society for Industrial and Applied Mathematics. <a href=\"https://doi.org/10.1137/1.9781611976465.38\">https://doi.org/10.1137/1.9781611976465.38</a>","ama":"Avni G, Jecker IR, Zikelic D. Infinite-duration all-pay bidding games. In: Marx D, ed. <i>Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms</i>. Society for Industrial and Applied Mathematics; 2021:617-636. doi:<a href=\"https://doi.org/10.1137/1.9781611976465.38\">10.1137/1.9781611976465.38</a>","short":"G. Avni, I.R. Jecker, D. Zikelic, in:, D. Marx (Ed.), Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms, Society for Industrial and Applied Mathematics, 2021, pp. 617–636."},"editor":[{"first_name":"Dániel","full_name":"Marx, Dániel","last_name":"Marx"}],"ec_funded":1,"status":"public","abstract":[{"text":"In a two-player zero-sum graph game the players move a token throughout a graph to produce an infinite path, which determines the winner or payoff of the game. Traditionally, the players alternate turns in moving the token. In bidding games, however, the players have budgets, and in each turn, we hold an “auction” (bidding) to determine which player moves the token: both players simultaneously submit bids and the higher bidder moves the token. The bidding mechanisms differ in their payment schemes. Bidding games were largely studied with variants of first-price bidding in which only the higher bidder pays his bid. We focus on all-pay bidding, where both players pay their bids. Finite-duration all-pay bidding games were studied and shown to be technically more challenging than their first-price counterparts. We study for the first time, infinite-duration all-pay bidding games. Our most interesting results are for mean-payoff objectives: we portray a complete picture for games played on strongly-connected graphs. We study both pure (deterministic) and mixed (probabilistic) strategies and completely characterize the optimal and almost-sure (with probability 1) payoffs the players can respectively guarantee. We show that mean-payoff games under all-pay bidding exhibit the intriguing mathematical properties of their first-price counterparts; namely, an equivalence with random-turn games in which in each turn, the player who moves is selected according to a (biased) coin toss. The equivalences for all-pay bidding are more intricate and unexpected than for first-price bidding.","lang":"eng"}],"conference":{"start_date":"2021-01-10","name":"SODA: Symposium on Discrete Algorithms","end_date":"2021-01-13","location":"Virtual"},"external_id":{"arxiv":["2005.06636"]},"oa_version":"Preprint","date_published":"2021-01-01T00:00:00Z","publication":"Proceedings of the 2021 ACM-SIAM Symposium on Discrete Algorithms","author":[{"id":"463C8BC2-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0001-5588-8287","last_name":"Avni","full_name":"Avni, Guy","first_name":"Guy"},{"id":"85D7C63E-7D5D-11E9-9C0F-98C4E5697425","full_name":"Jecker, Ismael R","first_name":"Ismael R","last_name":"Jecker"},{"orcid":"0000-0002-4681-1699","id":"294AA7A6-F248-11E8-B48F-1D18A9856A87","full_name":"Zikelic, Dorde","first_name":"Dorde","last_name":"Zikelic"}],"doi":"10.1137/1.9781611976465.38","publisher":"Society for Industrial and Applied Mathematics","language":[{"iso":"eng"}],"arxiv":1,"type":"conference","article_processing_charge":"No","publication_identifier":{"isbn":["978-1-61197-646-5"]},"main_file_link":[{"url":"https://arxiv.org/abs/2005.06636","open_access":"1"}],"scopus_import":"1","date_created":"2022-01-27T12:11:23Z","year":"2021","quality_controlled":"1","page":"617-636","project":[{"name":"The Wittgenstein Prize","grant_number":"Z211","_id":"25F42A32-B435-11E9-9278-68D0E5697425","call_identifier":"FWF"},{"_id":"0599E47C-7A3F-11EA-A408-12923DDC885E","grant_number":"863818","name":"Formal Methods for Stochastic Models: Algorithms and Applications","call_identifier":"H2020"},{"_id":"2564DBCA-B435-11E9-9278-68D0E5697425","name":"International IST Doctoral Program","grant_number":"665385","call_identifier":"H2020"}],"date_updated":"2025-07-14T09:10:12Z"},{"oa_version":"Published Version","conference":{"name":" ICLR: International Conference on Learning Representations","start_date":"2021-05-03","location":"Virtual","end_date":"2021-05-07"},"department":[{"_id":"GradSch"},{"_id":"ChLa"}],"date_published":"2021-05-01T00:00:00Z","publication":"9th International Conference on Learning Representations","author":[{"last_name":"Bui Thi Mai","first_name":"Phuong","full_name":"Bui Thi Mai, Phuong","id":"3EC6EE64-F248-11E8-B48F-1D18A9856A87"},{"orcid":"0000-0001-8622-7887","id":"40C20FD2-F248-11E8-B48F-1D18A9856A87","first_name":"Christoph","full_name":"Lampert, Christoph","last_name":"Lampert"}],"ddc":["000"],"_id":"9416","month":"05","user_id":"2DF688A6-F248-11E8-B48F-1D18A9856A87","language":[{"iso":"eng"}],"publication_status":"published","oa":1,"title":"The inductive bias of ReLU networks on orthogonally separable data","related_material":{"record":[{"id":"9418","status":"public","relation":"dissertation_contains"}]},"type":"conference","file_date_updated":"2021-05-24T11:15:57Z","day":"01","has_accepted_license":"1","article_processing_charge":"No","citation":{"mla":"Phuong, Mary, and Christoph Lampert. “The Inductive Bias of ReLU Networks on Orthogonally Separable Data.” <i>9th International Conference on Learning Representations</i>, 2021.","ieee":"M. Phuong and C. Lampert, “The inductive bias of ReLU networks on orthogonally separable data,” in <i>9th International Conference on Learning Representations</i>, Virtual, 2021.","ista":"Phuong M, Lampert C. 2021. The inductive bias of ReLU networks on orthogonally separable data. 9th International Conference on Learning Representations.  ICLR: International Conference on Learning Representations.","chicago":"Phuong, Mary, and Christoph Lampert. “The Inductive Bias of ReLU Networks on Orthogonally Separable Data.” In <i>9th International Conference on Learning Representations</i>, 2021.","apa":"Phuong, M., &#38; Lampert, C. (2021). The inductive bias of ReLU networks on orthogonally separable data. In <i>9th International Conference on Learning Representations</i>. Virtual.","ama":"Phuong M, Lampert C. The inductive bias of ReLU networks on orthogonally separable data. In: <i>9th International Conference on Learning Representations</i>. ; 2021.","short":"M. Phuong, C. Lampert, in:, 9th International Conference on Learning Representations, 2021."},"main_file_link":[{"url":"https://openreview.net/pdf?id=krz7T0xU9Z_","open_access":"1"}],"scopus_import":"1","year":"2021","date_created":"2021-05-24T11:16:46Z","quality_controlled":"1","date_updated":"2023-09-07T13:29:50Z","status":"public","file":[{"access_level":"open_access","date_updated":"2021-05-24T11:15:57Z","content_type":"application/pdf","file_id":"9417","date_created":"2021-05-24T11:15:57Z","file_name":"iclr2021_conference.pdf","creator":"bphuong","file_size":502356,"checksum":"f34ff17017527db5ba6927f817bdd125","relation":"main_file"}],"abstract":[{"lang":"eng","text":"We study the inductive bias of two-layer ReLU networks trained by gradient flow. We identify a class of easy-to-learn (`orthogonally separable') datasets, and characterise the solution that ReLU networks trained on such datasets converge to. Irrespective of network width, the solution turns out to be a combination of two max-margin classifiers: one corresponding to the positive data subset and one corresponding to the negative data subset. The proof is based on the recently introduced concept of extremal sectors, for which we prove a number of properties in the context of orthogonal separability. In particular, we prove stationarity of activation patterns from some time  onwards, which enables a reduction of the ReLU network to an ensemble of linear subnetworks."}]},{"doi":"10.15479/AT:ISTA:9418","publisher":"Institute of Science and Technology Austria","language":[{"iso":"eng"}],"type":"dissertation","file_date_updated":"2021-05-24T11:56:02Z","oa_version":"Published Version","date_published":"2021-05-30T00:00:00Z","author":[{"first_name":"Phuong","full_name":"Bui Thi Mai, Phuong","last_name":"Bui Thi Mai","id":"3EC6EE64-F248-11E8-B48F-1D18A9856A87"}],"ddc":["000"],"date_created":"2021-05-24T13:06:23Z","year":"2021","page":"125","acknowledged_ssus":[{"_id":"ScienComp"},{"_id":"CampIT"},{"_id":"E-Lib"}],"date_updated":"2023-09-08T11:11:12Z","file":[{"checksum":"4f0abe64114cfed264f9d36e8d1197e3","relation":"main_file","creator":"bphuong","file_size":2673905,"file_name":"mph-thesis-v519-pdfimages.pdf","date_updated":"2021-05-24T11:22:29Z","success":1,"access_level":"open_access","date_created":"2021-05-24T11:22:29Z","file_id":"9419","content_type":"application/pdf"},{"creator":"bphuong","file_size":92995100,"checksum":"f5699e876bc770a9b0df8345a77720a2","relation":"source_file","access_level":"closed","date_updated":"2021-05-24T11:56:02Z","file_id":"9420","content_type":"application/zip","date_created":"2021-05-24T11:56:02Z","file_name":"thesis.zip"}],"article_processing_charge":"No","publication_identifier":{"issn":["2663-337X"]},"month":"05","_id":"9418","user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","publication_status":"published","title":"Underspecification in deep learning","oa":1,"related_material":{"record":[{"relation":"part_of_dissertation","status":"deleted","id":"7435"},{"id":"7481","relation":"part_of_dissertation","status":"public"},{"id":"9416","relation":"part_of_dissertation","status":"public"},{"id":"7479","relation":"part_of_dissertation","status":"public"}]},"department":[{"_id":"GradSch"},{"_id":"ChLa"}],"alternative_title":["ISTA Thesis"],"degree_awarded":"PhD","status":"public","abstract":[{"lang":"eng","text":"Deep learning is best known for its empirical success across a wide range of applications\r\nspanning computer vision, natural language processing and speech. Of equal significance,\r\nthough perhaps less known, are its ramifications for learning theory: deep networks have\r\nbeen observed to perform surprisingly well in the high-capacity regime, aka the overfitting\r\nor underspecified regime. Classically, this regime on the far right of the bias-variance curve\r\nis associated with poor generalisation; however, recent experiments with deep networks\r\nchallenge this view.\r\n\r\nThis thesis is devoted to investigating various aspects of underspecification in deep learning.\r\nFirst, we argue that deep learning models are underspecified on two levels: a) any given\r\ntraining dataset can be fit by many different functions, and b) any given function can be\r\nexpressed by many different parameter configurations. We refer to the second kind of\r\nunderspecification as parameterisation redundancy and we precisely characterise its extent.\r\nSecond, we characterise the implicit criteria (the inductive bias) that guide learning in the\r\nunderspecified regime. Specifically, we consider a nonlinear but tractable classification\r\nsetting, and show that given the choice, neural networks learn classifiers with a large margin.\r\nThird, we consider learning scenarios where the inductive bias is not by itself sufficient to\r\ndeal with underspecification. We then study different ways of ‘tightening the specification’: i)\r\nIn the setting of representation learning with variational autoencoders, we propose a hand-\r\ncrafted regulariser based on mutual information. ii) In the setting of binary classification, we\r\nconsider soft-label (real-valued) supervision. We derive a generalisation bound for linear\r\nnetworks supervised in this way and verify that soft labels facilitate fast learning. Finally, we\r\nexplore an application of soft-label supervision to the training of multi-exit models."}],"day":"30","has_accepted_license":"1","supervisor":[{"id":"40C20FD2-F248-11E8-B48F-1D18A9856A87","orcid":"0000-0001-8622-7887","last_name":"Lampert","first_name":"Christoph","full_name":"Lampert, Christoph"}],"citation":{"ieee":"M. Phuong, “Underspecification in deep learning,” Institute of Science and Technology Austria, 2021.","mla":"Phuong, Mary. <i>Underspecification in Deep Learning</i>. Institute of Science and Technology Austria, 2021, doi:<a href=\"https://doi.org/10.15479/AT:ISTA:9418\">10.15479/AT:ISTA:9418</a>.","ista":"Phuong M. 2021. Underspecification in deep learning. Institute of Science and Technology Austria.","chicago":"Phuong, Mary. “Underspecification in Deep Learning.” Institute of Science and Technology Austria, 2021. <a href=\"https://doi.org/10.15479/AT:ISTA:9418\">https://doi.org/10.15479/AT:ISTA:9418</a>.","apa":"Phuong, M. (2021). <i>Underspecification in deep learning</i>. Institute of Science and Technology Austria. <a href=\"https://doi.org/10.15479/AT:ISTA:9418\">https://doi.org/10.15479/AT:ISTA:9418</a>","ama":"Phuong M. Underspecification in deep learning. 2021. doi:<a href=\"https://doi.org/10.15479/AT:ISTA:9418\">10.15479/AT:ISTA:9418</a>","short":"M. Phuong, Underspecification in Deep Learning, Institute of Science and Technology Austria, 2021."}}]