@phdthesis{9733, abstract = {This thesis is the result of the research carried out by the author during his PhD at IST Austria between 2017 and 2021. It mainly focuses on the Fröhlich polaron model, specifically to its regime of strong coupling. This model, which is rigorously introduced and discussed in the introduction, has been of great interest in condensed matter physics and field theory for more than eighty years. It is used to describe an electron interacting with the atoms of a solid material (the strength of this interaction is modeled by the presence of a coupling constant α in the Hamiltonian of the system). The particular regime examined here, which is mathematically described by considering the limit α →∞, displays many interesting features related to the emergence of classical behavior, which allows for a simplified effective description of the system under analysis. The properties, the range of validity and a quantitative analysis of the precision of such classical approximations are the main object of the present work. We specify our investigation to the study of the ground state energy of the system, its dynamics and its effective mass. For each of these problems, we provide in the introduction an overview of the previously known results and a detailed account of the original contributions by the author.}, author = {Feliciangeli, Dario}, issn = {2663-337X}, pages = {180}, publisher = {Institute of Science and Technology Austria}, title = {{The polaron at strong coupling}}, doi = {10.15479/at:ista:9733}, 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}, publisher = {IEEE}, 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}, issn = {1049-5258}, location = {Virtual}, publisher = {Neural Information Processing Systems Foundation}, title = {{Infinite time horizon safety of Bayesian neural networks}}, doi = {10.48550/arXiv.2111.03165}, 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}, issn = {1049-5258}, location = {Virtual}, publisher = {Neural Information Processing Systems Foundation}, title = {{Causal navigation by continuous-time neural networks}}, 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}, title = {{The generalized spatial representation in the prefrontal cortex is inherited from the hippocampus}}, doi = {10.1101/2021.09.30.462269}, year = {2021}, } @phdthesis{9418, abstract = {Deep learning is best known for its empirical success across a wide range of applications spanning computer vision, natural language processing and speech. Of equal significance, though perhaps less known, are its ramifications for learning theory: deep networks have been observed to perform surprisingly well in the high-capacity regime, aka the overfitting or underspecified regime. Classically, this regime on the far right of the bias-variance curve is associated with poor generalisation; however, recent experiments with deep networks challenge this view. This thesis is devoted to investigating various aspects of underspecification in deep learning. First, we argue that deep learning models are underspecified on two levels: a) any given training dataset can be fit by many different functions, and b) any given function can be expressed by many different parameter configurations. We refer to the second kind of underspecification as parameterisation redundancy and we precisely characterise its extent. Second, we characterise the implicit criteria (the inductive bias) that guide learning in the underspecified regime. Specifically, we consider a nonlinear but tractable classification setting, and show that given the choice, neural networks learn classifiers with a large margin. Third, we consider learning scenarios where the inductive bias is not by itself sufficient to deal with underspecification. We then study different ways of ‘tightening the specification’: i) In the setting of representation learning with variational autoencoders, we propose a hand- crafted regulariser based on mutual information. ii) In the setting of binary classification, we consider soft-label (real-valued) supervision. We derive a generalisation bound for linear networks supervised in this way and verify that soft labels facilitate fast learning. Finally, we explore an application of soft-label supervision to the training of multi-exit models.}, author = {Bui Thi Mai, Phuong}, issn = {2663-337X}, pages = {125}, publisher = {Institute of Science and Technology Austria}, title = {{Underspecification in deep learning}}, doi = {10.15479/AT:ISTA:9418}, year = {2021}, } @inproceedings{9416, abstract = {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.}, author = {Bui Thi Mai, Phuong and Lampert, Christoph}, booktitle = {9th International Conference on Learning Representations}, location = {Virtual}, title = {{The inductive bias of ReLU networks on orthogonally separable data}}, year = {2021}, } @inproceedings{9356, abstract = {In runtime verification, a monitor watches a trace of a system and, if possible, decides after observing each finite prefix whether or not the unknown infinite trace satisfies a given specification. We generalize the theory of runtime verification to monitors that attempt to estimate numerical values of quantitative trace properties (instead of attempting to conclude boolean values of trace specifications), such as maximal or average response time along a trace. Quantitative monitors are approximate: with every finite prefix, they can improve their estimate of the infinite trace's unknown property value. Consequently, quantitative monitors can be compared with regard to a precision-cost trade-off: better approximations of the property value require more monitor resources, such as states (in the case of finite-state monitors) or registers, and additional resources yield better approximations. We introduce a formal framework for quantitative and approximate monitoring, show how it conservatively generalizes the classical boolean setting for monitoring, and give several precision-cost trade-offs for monitors. For example, we prove that there are quantitative properties for which every additional register improves monitoring precision.}, author = {Henzinger, Thomas A and Sarac, Naci E}, booktitle = {Proceedings of the 36th Annual ACM/IEEE Symposium on Logic in Computer Science}, location = {Online}, publisher = {IEEE}, title = {{Quantitative and approximate monitoring}}, doi = {10.1109/LICS52264.2021.9470547}, 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}}, doi = {10.1609/aaai.v35i5.16496}, volume = {35}, year = {2021}, } @inproceedings{10671, abstract = {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.}, author = {Hasani, Ramin and Lechner, Mathias and Amini, Alexander and Rus, Daniela and Grosu, Radu}, booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence}, isbn = {978-1-57735-866-4}, issn = {2374-3468}, location = {Virtual}, number = {9}, pages = {7657--7666}, publisher = {AAAI Press}, title = {{Liquid time-constant networks}}, doi = {10.1609/aaai.v35i9.16936}, volume = {35}, 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}}, doi = {10.1609/aaai.v35i13.17372}, volume = {35}, 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{9558, abstract = {We show that turbulent dynamics that arise in simulations of the three-dimensional Navier--Stokes equations in a triply-periodic domain under sinusoidal forcing can be described as transient visits to the neighborhoods of unstable time-periodic solutions. Based on this description, we reduce the original system with more than 10^5 degrees of freedom to a 17-node Markov chain where each node corresponds to the neighborhood of a periodic orbit. The model accurately reproduces long-term averages of the system's observables as weighted sums over the periodic orbits. }, author = {Yalniz, Gökhan and Hof, Björn and Budanur, Nazmi B}, issn = {1079-7114}, journal = {Physical Review Letters}, number = {24}, publisher = {American Physical Society}, title = {{Coarse graining the state space of a turbulent flow using periodic orbits}}, doi = {10.1103/PhysRevLett.126.244502}, volume = {126}, year = {2021}, } @phdthesis{10035, abstract = {Many security definitions come in two flavors: a stronger “adaptive” flavor, where the adversary can arbitrarily make various choices during the course of the attack, and a weaker “selective” flavor where the adversary must commit to some or all of their choices a-priori. For example, in the context of identity-based encryption, selective security requires the adversary to decide on the identity of the attacked party at the very beginning of the game whereas adaptive security allows the attacker to first see the master public key and some secret keys before making this choice. Often, it appears to be much easier to achieve selective security than it is to achieve adaptive security. A series of several recent works shows how to cleverly achieve adaptive security in several such scenarios including generalized selective decryption [Pan07][FJP15], constrained PRFs [FKPR14], and Yao’s garbled circuits [JW16]. Although the above works expressed vague intuition that they share a common technique, the connection was never made precise. In this work we present a new framework (published at Crypto ’17 [JKK+17a]) that connects all of these works and allows us to present them in a unified and simplified fashion. Having the framework in place, we show how to achieve adaptive security for proxy re-encryption schemes (published at PKC ’19 [FKKP19]) and provide the first adaptive security proofs for continuous group key agreement protocols (published at S&P ’21 [KPW+21]). Questioning optimality of our framework, we then show that currently used proof techniques cannot lead to significantly better security guarantees for "graph-building" games (published at TCC ’21 [KKPW21a]). These games cover generalized selective decryption, as well as the security of prominent constructions for constrained PRFs, continuous group key agreement, and proxy re-encryption. Finally, we revisit the adaptive security of Yao’s garbled circuits and extend the analysis of Jafargholi and Wichs in two directions: While they prove adaptive security only for a modified construction with increased online complexity, we provide the first positive results for the original construction by Yao (published at TCC ’21 [KKP21a]). On the negative side, we prove that the results of Jafargholi and Wichs are essentially optimal by showing that no black-box reduction can provide a significantly better security bound (published at Crypto ’21 [KKPW21c]).}, author = {Klein, Karen}, issn = {2663-337X}, pages = {276}, publisher = {Institute of Science and Technology Austria}, title = {{On the adaptive security of graph-based games}}, doi = {10.15479/at:ista:10035}, year = {2021}, } @inproceedings{9826, abstract = {Automated contract tracing aims at supporting manual contact tracing during pandemics by alerting users of encounters with infected people. There are currently many proposals for protocols (like the “decentralized” DP-3T and PACT or the “centralized” ROBERT and DESIRE) to be run on mobile phones, where the basic idea is to regularly broadcast (using low energy Bluetooth) some values, and at the same time store (a function of) incoming messages broadcasted by users in their proximity. In the existing proposals one can trigger false positives on a massive scale by an “inverse-Sybil” attack, where a large number of devices (malicious users or hacked phones) pretend to be the same user, such that later, just a single person needs to be diagnosed (and allowed to upload) to trigger an alert for all users who were in proximity to any of this large group of devices. We propose the first protocols that do not succumb to such attacks assuming the devices involved in the attack do not constantly communicate, which we observe is a necessary assumption. The high level idea of the protocols is to derive the values to be broadcasted by a hash chain, so that two (or more) devices who want to launch an inverse-Sybil attack will not be able to connect their respective chains and thus only one of them will be able to upload. Our protocols also achieve security against replay, belated replay, and one of them even against relay attacks.}, author = {Auerbach, Benedikt and Chakraborty, Suvradip and Klein, Karen and Pascual Perez, Guillermo and Pietrzak, Krzysztof Z and Walter, Michael and Yeo, Michelle X}, booktitle = {Topics in Cryptology – CT-RSA 2021}, isbn = {9783030755386}, issn = {1611-3349}, location = {Virtual Event}, pages = {399--421}, publisher = {Springer Nature}, title = {{Inverse-Sybil attacks in automated contact tracing}}, doi = {10.1007/978-3-030-75539-3_17}, volume = {12704}, year = {2021}, } @phdthesis{9397, abstract = {Accumulation of interstitial fluid (IF) between embryonic cells is a common phenomenon in vertebrate embryogenesis. Unlike other model systems, where these accumulations coalesce into a large central cavity – the blastocoel, in zebrafish, IF is more uniformly distributed between the deep cells (DC) before the onset of gastrulation. This is likely due to the presence of a large extraembryonic structure – the yolk cell (YC) at the position where the blastocoel typically forms in other model organisms. IF has long been speculated to play a role in tissue morphogenesis during embryogenesis, but direct evidence supporting such function is still sparse. Here we show that the relocalization of IF to the interface between the YC and DC/epiblast is critical for axial mesendoderm (ME) cell protrusion formation and migration along this interface, a key process in embryonic axis formation. We further demonstrate that axial ME cell migration and IF relocalization engage in a positive feedback loop, where axial ME migration triggers IF accumulation ahead of the advancing axial ME tissue by mechanically compressing the overlying epiblast cell layer. Upon compression, locally induced flow relocalizes the IF through the porous epiblast tissue resulting in an IF accumulation ahead of the leading axial ME. This IF accumulation, in turn, promotes cell protrusion formation and migration of the leading axial ME cells, thereby facilitating axial ME extension. Our findings reveal a central role of dynamic IF relocalization in orchestrating germ layer morphogenesis during gastrulation.}, author = {Huljev, Karla}, issn = {2663-337X}, pages = {101}, publisher = {Institute of Science and Technology Austria}, title = {{Coordinated spatiotemporal reorganization of interstitial fluid is required for axial mesendoderm migration in zebrafish gastrulation}}, doi = {10.15479/at:ista:9397}, year = {2021}, } @phdthesis{9992, abstract = {Blood – this is what animals use to heal wounds fast and efficient. Plants do not have blood circulation and their cells cannot move. However, plants have evolved remarkable capacities to regenerate tissues and organs preventing further damage. In my PhD research, I studied the wound healing in the Arabidopsis root. I used a UV laser to ablate single cells in the root tip and observed the consequent wound healing. Interestingly, the inner adjacent cells induced a division plane switch and subsequently adopted the cell type of the killed cell to replace it. We termed this form of wound healing “restorative divisions”. This initial observation triggered the questions of my PhD studies: How and why do cells orient their division planes, how do they feel the wound and why does this happen only in inner adjacent cells. For answering these questions, I used a quite simple experimental setup: 5 day - old seedlings were stained with propidium iodide to visualize cell walls and dead cells; ablation was carried out using a special laser cutter and a confocal microscope. Adaptation of the novel vertical microscope system made it possible to observe wounds in real time. This revealed that restorative divisions occur at increased frequency compared to normal divisions. Additionally, the major plant hormone auxin accumulates in wound adjacent cells and drives the expression of the wound-stress responsive transcription factor ERF115. Using this as a marker gene for wound responses, we found that an important part of wound signalling is the sensing of the collapse of the ablated cell. The collapse causes a radical pressure drop, which results in strong tissue deformations. These deformations manifest in an invasion of the now free spot specifically by the inner adjacent cells within seconds, probably because of higher pressure of the inner tissues. Long-term imaging revealed that those deformed cells continuously expand towards the wound hole and that this is crucial for the restorative division. These wound-expanding cells exhibit an abnormal, biphasic polarity of microtubule arrays before the division. Experiments inhibiting cell expansion suggest that it is the biphasic stretching that induces those MT arrays. Adapting the micromanipulator aspiration system from animal scientists at our institute confirmed the hypothesis that stretching influences microtubule stability. In conclusion, this shows that microtubules react to tissue deformation and this facilitates the observed division plane switch. This puts mechanical cues and tensions at the most prominent position for explaining the growth and wound healing properties of plants. Hence, it shines light onto the importance of understanding mechanical signal transduction. }, author = {Hörmayer, Lukas}, issn = {2663-337X}, pages = {168}, publisher = {Institute of Science and Technology Austria}, title = {{Wound healing in the Arabidopsis root meristem}}, doi = {10.15479/at:ista:9992}, year = {2021}, } @phdthesis{9728, abstract = {Most real-world flows are multiphase, yet we know little about them compared to their single-phase counterparts. Multiphase flows are more difficult to investigate as their dynamics occur in large parameter space and involve complex phenomena such as preferential concentration, turbulence modulation, non-Newtonian rheology, etc. Over the last few decades, experiments in particle-laden flows have taken a back seat in favour of ever-improving computational resources. However, computers are still not powerful enough to simulate a real-world fluid with millions of finite-size particles. Experiments are essential not only because they offer a reliable way to investigate real-world multiphase flows but also because they serve to validate numerical studies and steer the research in a relevant direction. In this work, we have experimentally investigated particle-laden flows in pipes, and in particular, examined the effect of particles on the laminar-turbulent transition and the drag scaling in turbulent flows. For particle-laden pipe flows, an earlier study [Matas et al., 2003] reported how the sub-critical (i.e., hysteretic) transition that occurs via localised turbulent structures called puffs is affected by the addition of particles. In this study, in addition to this known transition, we found a super-critical transition to a globally fluctuating state with increasing particle concentration. At the same time, the Newtonian-type transition via puffs is delayed to larger Reynolds numbers. At an even higher concentration, only the globally fluctuating state is found. The dynamics of particle-laden flows are hence determined by two competing instabilities that give rise to three flow regimes: Newtonian-type turbulence at low, a particle-induced globally fluctuating state at high, and a coexistence state at intermediate concentrations. The effect of particles on turbulent drag is ambiguous, with studies reporting drag reduction, no net change, and even drag increase. The ambiguity arises because, in addition to particle concentration, particle shape, size, and density also affect the net drag. Even similar particles might affect the flow dissimilarly in different Reynolds number and concentration ranges. In the present study, we explored a wide range of both Reynolds number and concentration, using spherical as well as cylindrical particles. We found that the spherical particles do not reduce drag while the cylindrical particles are drag-reducing within a specific Reynolds number interval. The interval strongly depends on the particle concentration and the relative size of the pipe and particles. Within this interval, the magnitude of drag reduction reaches a maximum. These drag reduction maxima appear to fall onto a distinct power-law curve irrespective of the pipe diameter and particle concentration, and this curve can be considered as the maximum drag reduction asymptote for a given fibre shape. Such an asymptote is well known for polymeric flows but had not been identified for particle-laden flows prior to this work.}, author = {Agrawal, Nishchal}, issn = {2663-337X}, keywords = {Drag Reduction, Transition to Turbulence, Multiphase Flows, particle Laden Flows, Complex Flows, Experiments, Fluid Dynamics}, pages = {118}, publisher = {Institute of Science and Technology Austria}, title = {{Transition to turbulence and drag reduction in particle-laden pipe flows}}, doi = {10.15479/at:ista:9728}, 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}, title = {{The structure of hippocampal CA1 interactions optimizes spatial coding across experience}}, doi = {10.1101/2021.09.28.460602}, year = {2021}, } @phdthesis{9623, abstract = {Cytoplasmic reorganizations are essential for morphogenesis. In large cells like oocytes, these reorganizations become crucial in patterning the oocyte for later stages of embryonic development. Ascidians oocytes reorganize their cytoplasm (ooplasm) in a spectacular manner. Ooplasmic reorganization is initiated at fertilization with the contraction of the actomyosin cortex along the animal-vegetal axis of the oocyte, driving the accumulation of cortical endoplasmic reticulum (cER), maternal mRNAs associated to it and a mitochondria-rich subcortical layer – the myoplasm – in a region of the vegetal pole termed contraction pole (CP). Here we have used the species Phallusia mammillata to investigate the changes in cell shape that accompany these reorganizations and the mechanochemical mechanisms underlining CP formation. We report that the length of the animal-vegetal (AV) axis oscillates upon fertilization: it first undergoes a cycle of fast elongation-lengthening followed by a slow expansion of mainly the vegetal pole (VP) of the cell. We show that the fast oscillation corresponds to a dynamic polarization of the actin cortex as a result of a fertilization-induced increase in cortical tension in the oocyte that triggers a rupture of the cortex at the animal pole and the establishment of vegetal-directed cortical flows. These flows are responsible for the vegetal accumulation of actin causing the VP to flatten. We find that the slow expansion of the VP, leading to CP formation, correlates with a relaxation of the vegetal cortex and that the myoplasm plays a role in the expansion. We show that the myoplasm is a solid-like layer that buckles under compression forces arising from the contracting actin cortex at the VP. Straightening of the myoplasm when actin flows stops, facilitates the expansion of the VP and the CP. Altogether, our results present a previously unrecognized role for the myoplasm in ascidian ooplasmic segregation. }, author = {Caballero Mancebo, Silvia}, isbn = {978-3-99078-012-1}, issn = {2663-337X}, pages = {111}, publisher = {Institute of Science and Technology Austria}, title = {{Fertilization-induced deformations are controlled by the actin cortex and a mitochondria-rich subcortical layer in ascidian oocytes}}, doi = {10.15479/at:ista:9623}, year = {2021}, }