@unpublished{10045, abstract = {Given a fixed finite metric space (V,μ), the {\em minimum 0-extension problem}, denoted as 0-Ext[μ], is equivalent to the following optimization problem: minimize function of the form minx∈Vn∑ifi(xi)+∑ijcijμ(xi,xj) where cij,cvi are given nonnegative costs and fi:V→R are functions given by fi(xi)=∑v∈Vcviμ(xi,v). The computational complexity of 0-Ext[μ] has been recently established by Karzanov and by Hirai: if metric μ is {\em orientable modular} then 0-Ext[μ] can be solved in polynomial time, otherwise 0-Ext[μ] is NP-hard. To prove the tractability part, Hirai developed a theory of discrete convex functions on orientable modular graphs generalizing several known classes of functions in discrete convex analysis, such as L♮-convex functions. We consider a more general version of the problem in which unary functions fi(xi) can additionally have terms of the form cuv;iμ(xi,{u,v}) for {u,v}∈F, where set F⊆(V2) is fixed. We extend the complexity classification above by providing an explicit condition on (μ,F) for the problem to be tractable. In order to prove the tractability part, we generalize Hirai's theory and define a larger class of discrete convex functions. It covers, in particular, another well-known class of functions, namely submodular functions on an integer lattice. Finally, we improve the complexity of Hirai's algorithm for solving 0-Ext on orientable modular graphs. }, author = {Dvorak, Martin and Kolmogorov, Vladimir}, booktitle = {arXiv}, keywords = {minimum 0-extension problem, metric labeling problem, discrete metric spaces, metric extensions, computational complexity, valued constraint satisfaction problems, discrete convex analysis, L-convex functions}, title = {{Generalized minimum 0-extension problem and discrete convexity}}, doi = {10.48550/arXiv.2109.10203}, year = {2021}, } @phdthesis{10058, abstract = {Quantum information and computation has become a vast field paved with opportunities for researchers and investors. As large multinational companies and international funds are heavily investing in quantum technologies it is still a question which platform is best suited for the task of realizing a scalable quantum processor. In this work we investigate hole spins in Ge quantum wells. These hold great promise as they possess several favorable properties: a small effective mass, a strong spin-orbit coupling, long relaxation time and an inherent immunity to hyperfine noise. All these characteristics helped Ge hole spin qubits to evolve from a single qubit to a fully entangled four qubit processor in only 3 years. Here, we investigated a qubit approach leveraging the large out-of-plane g-factors of heavy hole states in Ge quantum dots. We found this qubit to be reproducibly operable at extremely low magnetic field and at large speeds while maintaining coherence. This was possible because large differences of g-factors in adjacent dots can be achieved in the out-of-plane direction. In the in-plane direction the small g-factors, on the other hand, can be altered very effectively by the confinement potentials. Here, we found that this can even lead to a sign change of the g-factors. The resulting g-factor difference alters the dynamics of the system drastically and produces effects typically attributed to a spin-orbit induced spin-flip term. The investigations carried out in this thesis give further insights into the possibilities of holes in Ge and reveal new physical properties that need to be considered when designing future spin qubit experiments.}, author = {Jirovec, Daniel}, issn = {2663-337X}, keywords = {qubits, quantum computing, holes}, pages = {151}, publisher = {Institute of Science and Technology Austria}, title = {{Singlet-Triplet qubits and spin-orbit interaction in 2-dimensional Ge hole gases}}, doi = {10.15479/at:ista:10058}, year = {2021}, } @unpublished{10077, abstract = {Although much is known about how single neurons in the hippocampus represent an animal’s position, how cell-cell interactions contribute to spatial coding remains poorly understood. Using a novel statistical estimator and theoretical modeling, both developed in the framework of maximum entropy models, we reveal highly structured cell-to-cell interactions whose statistics depend on familiar vs. novel environment. In both conditions the circuit interactions optimize the encoding of spatial information, but for regimes that differ in the signal-to-noise ratio of their spatial inputs. Moreover, the topology of the interactions facilitates linear decodability, making the information easy to read out by downstream circuits. These findings suggest that the efficient coding hypothesis is not applicable only to individual neuron properties in the sensory periphery, but also to neural interactions in the central brain.}, author = {Nardin, Michele and Csicsvari, Jozsef L and Tkačik, Gašper and Savin, Cristina}, booktitle = {bioRxiv}, publisher = {Cold Spring Harbor Laboratory}, title = {{The structure of hippocampal CA1 interactions optimizes spatial coding across experience}}, doi = {10.1101/2021.09.28.460602}, year = {2021}, } @unpublished{10080, abstract = {Hippocampal and neocortical neural activity is modulated by the position of the individual in space. While hippocampal neurons provide the basis for a spatial map, prefrontal cortical neurons generalize over environmental features. Whether these generalized representations result from a bidirectional interaction with, or are mainly derived from hippocampal spatial representations is not known. By examining simultaneously recorded hippocampal and medial prefrontal neurons, we observed that prefrontal spatial representations show a delayed coherence with hippocampal ones. We also identified subpopulations of cells in the hippocampus and medial prefrontal cortex that formed functional cross-area couplings; these resembled the optimal connections predicted by a probabilistic model of spatial information transfer and generalization. Moreover, cross-area couplings were strongest and had the shortest delay preceding spatial decision-making. Our results suggest that generalized spatial coding in the medial prefrontal cortex is inherited from spatial representations in the hippocampus, and that the routing of information can change dynamically with behavioral demands.}, author = {Nardin, Michele and Käfer, Karola and Csicsvari, Jozsef L}, booktitle = {bioRxiv}, publisher = {Cold Spring Harbor Laboratory}, title = {{The generalized spatial representation in the prefrontal cortex is inherited from the hippocampus}}, doi = {10.1101/2021.09.30.462269}, year = {2021}, } @phdthesis{10083, abstract = {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 combined 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 alkalinization 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. Besides, transgenic analysis combined with genetic engineering and biochemistry showed that vi they 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 growth, novel auxin signaling pathway as well as auxin-peptide crosstalk. }, author = {Li, Lanxin}, issn = {2663-337X}, publisher = {Institute of Science and Technology Austria}, title = {{Rapid cell growth regulation in Arabidopsis}}, doi = {10.15479/at:ista:10083}, year = {2021}, } @phdthesis{10135, abstract = {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 cytokinin 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 translation. 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 chemical 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 cytokinin 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.}, author = {Semerádová, Hana}, isbn = {978-3-99078-014-5}, issn = {2663-337X}, publisher = {Institute of Science and Technology Austria}, title = {{Molecular mechanisms of the cytokinin-regulated endomembrane trafficking to coordinate plant organogenesis}}, doi = {10.15479/at:ista:10135}, year = {2021}, } @article{10191, abstract = {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. }, author = {Bui, Truc Lam and Chatterjee, Krishnendu and Gautam, Tushar and Pavlogiannis, Andreas and Toman, Viktor}, issn = {2475-1421}, journal = {Proceedings of the ACM on Programming Languages}, keywords = {safety, risk, reliability and quality, software}, number = {OOPSLA}, publisher = {Association for Computing Machinery}, title = {{The reads-from equivalence for the TSO and PSO memory models}}, doi = {10.1145/3485541}, volume = {5}, year = {2021}, } @phdthesis{10199, abstract = {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.}, author = {Toman, Viktor}, issn = {2663-337X}, keywords = {concurrency, verification, model checking}, pages = {166}, publisher = {Institute of Science and Technology Austria}, title = {{Improved verification techniques for concurrent systems}}, doi = {10.15479/at:ista:10199}, year = {2021}, } @phdthesis{10293, abstract = {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 than 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 are 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.}, author = {Schmid, Laura}, issn = {2663-337X}, pages = {171}, publisher = {Institute of Science and Technology Austria}, title = {{Evolution of cooperation via (in)direct reciprocity under imperfect information}}, doi = {10.15479/at:ista:10293}, year = {2021}, } @phdthesis{10303, abstract = {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. }, author = {Abualia, Rashed}, issn = {2663-337X}, pages = {139}, publisher = {Institute of Science and Technology Austria}, title = {{Role of hormones in nitrate regulated growth}}, doi = {10.15479/at:ista:10303}, year = {2021}, } @phdthesis{10307, abstract = {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.}, author = {Tomasek, Kathrin}, issn = {2663-337X}, pages = {73}, publisher = {Institute of Science and Technology Austria}, title = {{Pathogenic Escherichia coli hijack the host immune response}}, doi = {10.15479/at:ista:10307}, year = {2021}, } @phdthesis{10422, abstract = {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.}, author = {Piankov, Anton}, issn = {2791-4585}, publisher = {Institute of Science and Technology Austria}, title = {{Towards designer materials using customizable particle shape}}, doi = {10.15479/at:ista:10422}, year = {2021}, } @phdthesis{10429, abstract = {The scalability of concurrent data structures and distributed algorithms strongly depends on reducing 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 such 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.}, author = {Nadiradze, Giorgi}, issn = {2663-337X}, pages = {132}, publisher = {Institute of Science and Technology Austria}, title = {{On achieving scalability through relaxation}}, doi = {10.15479/at:ista:10429}, year = {2021}, } @article{10635, abstract = {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.}, author = {Nardin, Michele and Phillips, James W. and Podlaski, William F. and Keemink, Sander W.}, issn = {2804-3871}, journal = {Peer Community Journal}, publisher = {Centre Mersenne ; Peer Community In}, title = {{Nonlinear computations in spiking neural networks through multiplicative synapses}}, doi = {10.24072/pcjournal.69}, volume = {1}, year = {2021}, } @inproceedings{10665, abstract = {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 into 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.}, author = {Henzinger, Thomas A and Lechner, Mathias and Zikelic, Dorde}, booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence}, isbn = {978-1-57735-866-4}, issn = {2374-3468}, location = {Virtual}, number = {5A}, pages = {3787--3795}, publisher = {AAAI Press}, title = {{Scalable verification of quantized neural networks}}, volume = {35}, year = {2021}, } @inproceedings{10666, abstract = {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.}, author = {Lechner, Mathias and Hasani, Ramin and Grosu, Radu and Rus, Daniela and Henzinger, Thomas A}, booktitle = {2021 IEEE International Conference on Robotics and Automation}, isbn = {978-1-7281-9078-5}, issn = {2577-087X}, location = {Xi'an, China}, pages = {4140--4147}, title = {{Adversarial training is not ready for robot learning}}, doi = {10.1109/ICRA48506.2021.9561036}, year = {2021}, } @inproceedings{10667, abstract = {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.}, author = {Lechner, Mathias and Žikelić, Ðorđe and Chatterjee, Krishnendu and Henzinger, Thomas A}, booktitle = {35th Conference on Neural Information Processing Systems}, location = {Virtual}, title = {{Infinite time horizon safety of Bayesian neural networks}}, doi = {10.48550/arXiv.2111.03165}, year = {2021}, } @inproceedings{10668, abstract = {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.}, author = {Babaiee, Zahra and Hasani, Ramin and Lechner, Mathias and Rus, Daniela and Grosu, Radu}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, issn = {2640-3498}, location = {Virtual}, pages = {478--489}, publisher = {ML Research Press}, title = {{On-off center-surround receptive fields for accurate and robust image classification}}, volume = {139}, year = {2021}, } @inproceedings{10669, abstract = {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 abstraction-based technique for constructing a tight Reachtube (an over-approximation of the set of reachable states over 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.}, author = {Grunbacher, Sophie and Hasani, Ramin and Lechner, Mathias and Cyranka, Jacek and Smolka, Scott A and Grosu, Radu}, booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence}, isbn = {978-1-57735-866-4}, issn = {2374-3468}, location = {Virtual}, number = {13}, pages = {11525--11535}, publisher = {AAAI Press}, title = {{On the verification of neural ODEs with stochastic guarantees}}, volume = {35}, year = {2021}, } @inproceedings{10670, abstract = {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 deep 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.}, author = {Vorbach, Charles J and Hasani, Ramin and Amini, Alexander and Lechner, Mathias and Rus, Daniela}, booktitle = {35th Conference on Neural Information Processing Systems}, location = {Virtual}, title = {{Causal navigation by continuous-time neural networks}}, year = {2021}, }