TY - THES
AB - This thesis is based on three main topics: In the first part, we study convergence of discrete gradient flow structures associated with regular finite-volume discretisations of Fokker-Planck equations. We show evolutionary I convergence of the discrete gradient flows to the L2-Wasserstein gradient flow corresponding to the solution of a Fokker-Planck
equation in arbitrary dimension d >= 1. Along the argument, we prove Mosco- and I-convergence results for discrete energy functionals, which are of independent interest for convergence of equivalent gradient flow structures in Hilbert spaces.
The second part investigates L2-Wasserstein flows on metric graph. The starting point is a Benamou-Brenier formula for the L2-Wasserstein distance, which is proved via a regularisation scheme for solutions of the continuity equation, adapted to the peculiar geometric structure of metric graphs. Based on those results, we show that the L2-Wasserstein space over a metric graph admits a gradient flow which may be identified as a solution of a Fokker-Planck equation.
In the third part, we focus again on the discrete gradient flows, already encountered in the first part. We propose a variational structure which extends the gradient flow structure to Markov chains violating the detailed-balance conditions. Using this structure, we characterise contraction estimates for the discrete heat flow in terms of convexity of
corresponding path-dependent energy functionals. In addition, we use this approach to derive several functional inequalities for said functionals.
AU - Forkert, Dominik L
ID - 7629
SN - 2663-337X
TI - Gradient flows in spaces of probability measures for finite-volume schemes, metric graphs and non-reversible Markov chains
ER -
TY - JOUR
AB - The posterior parietal cortex (PPC) and frontal motor areas comprise a cortical network supporting goal-directed behaviour, with functions including sensorimotor transformations and decision making. In primates, this network links performed and observed actions via mirror neurons, which fire both when individuals perform an action and when they observe the same action performed by a conspecific. Mirror neurons are believed to be important for social learning, but it is not known whether mirror-like neurons occur in similar networks in other social species, such as rodents, or if they can be measured in such models using paradigms where observers passively view a demonstrator. Therefore, we imaged Ca2+ responses in PPC and secondary motor cortex (M2) while mice performed and observed pellet-reaching and wheel-running tasks, and found that cell populations in both areas robustly encoded several naturalistic behaviours. However, neural responses to the same set of observed actions were absent, although we verified that observer mice were attentive to performers and that PPC neurons responded reliably to visual cues. Statistical modelling also indicated that executed actions outperformed observed actions in predicting neural responses. These results raise the possibility that sensorimotor action recognition in rodents could take place outside of the parieto-frontal circuit, and underscore that detecting socially-driven neural coding depends critically on the species and behavioural paradigm used.
AU - Tombaz, Tuce
AU - Dunn, Benjamin A.
AU - Hovde, Karoline
AU - Cubero, Ryan J
AU - Mimica, Bartul
AU - Mamidanna, Pranav
AU - Roudi, Yasser
AU - Whitlock, Jonathan R.
ID - 7632
IS - 1
JF - Scientific reports
TI - Action representation in the mouse parieto-frontal network
VL - 10
ER -
TY - JOUR
AB - Assemblies of colloidal semiconductor nanocrystals (NCs) in the form of thin solid films leverage the size-dependent quantum confinement properties and the wet chemical methods vital for the development of the emerging solution-processable electronics, photonics, and optoelectronics technologies. The ability to control the charge carrier transport in the colloidal NC assemblies is fundamental for altering their electronic and optical properties for the desired applications. Here we demonstrate a strategy to render the solids of narrow-bandgap NC assemblies exclusively electron-transporting by creating a type-II heterojunction via shelling. Electronic transport of molecularly cross-linked PbTe@PbS core@shell NC assemblies is measured using both a conventional solid gate transistor and an electric-double-layer transistor, as well as compared with those of core-only PbTe NCs. In contrast to the ambipolar characteristics demonstrated by many narrow-bandgap NCs, the core@shell NCs exhibit exclusive n-type transport, i.e., drastically suppressed contribution of holes to the overall transport. The PbS shell that forms a type-II heterojunction assists the selective carrier transport by heavy doping of electrons into the PbTe-core conduction level and simultaneously strongly localizes the holes within the NC core valence level. This strongly enhanced n-type transport makes these core@shell NCs suitable for applications where ambipolar characteristics should be actively suppressed, in particular, for thermoelectric and electron-transporting layers in photovoltaic devices.
AU - Miranti, Retno
AU - Shin, Daiki
AU - Septianto, Ricky Dwi
AU - Ibáñez, Maria
AU - Kovalenko, Maksym V.
AU - Matsushita, Nobuhiro
AU - Iwasa, Yoshihiro
AU - Bisri, Satria Zulkarnaen
ID - 7634
IS - 3
JF - ACS Nano
TI - Exclusive electron transport in Core@Shell PbTe@PbS colloidal semiconductor nanocrystal assemblies
VL - 14
ER -
TY - CONF
AB - Concurrent programming can be notoriously complex and error-prone. Programming bugs can arise from a variety of sources, such as operation re-reordering, or incomplete understanding of the memory model. A variety of formal and model checking methods have been developed to address this fundamental difficulty. While technically interesting, existing academic methods are still hard to apply to the large codebases typical of industrial deployments, which limits their practical impact.
AU - Koval, Nikita
AU - Sokolova, Mariia
AU - Fedorov, Alexander
AU - Alistarh, Dan-Adrian
AU - Tsitelov, Dmitry
ID - 7635
SN - 9781450368186
T2 - Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP
TI - Testing concurrency on the JVM with Lincheck
ER -
TY - CONF
AB - Balanced search trees typically use key comparisons to guide their operations, and achieve logarithmic running time. By relying on numerical properties of the keys, interpolation search achieves lower search complexity and better performance. Although interpolation-based data structures were investigated in the past, their non-blocking concurrent variants have received very little attention so far.
In this paper, we propose the first non-blocking implementation of the classic interpolation search tree (IST) data structure. For arbitrary key distributions, the data structure ensures worst-case O(log n + p) amortized time for search, insertion and deletion traversals. When the input key distributions are smooth, lookups run in expected O(log log n + p) time, and insertion and deletion run in expected amortized O(log log n + p) time, where p is a bound on the number of threads. To improve the scalability of concurrent insertion and deletion, we propose a novel parallel rebuilding technique, which should be of independent interest.
We evaluate whether the theoretical improvements translate to practice by implementing the concurrent interpolation search tree, and benchmarking it on uniform and nonuniform key distributions, for dataset sizes in the millions to billions of keys. Relative to the state-of-the-art concurrent data structures, the concurrent interpolation search tree achieves performance improvements of up to 15% under high update rates, and of up to 50% under moderate update rates. Further, ISTs exhibit up to 2X less cache-misses, and consume 1.2 -- 2.6X less memory compared to the next best alternative on typical dataset sizes. We find that the results are surprisingly robust to distributional skew, which suggests that our data structure can be a promising alternative to classic concurrent search structures.
AU - Brown, Trevor A
AU - Prokopec, Aleksandar
AU - Alistarh, Dan-Adrian
ID - 7636
SN - 9781450368186
T2 - Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP
TI - Non-blocking interpolation search trees with doubly-logarithmic running time
ER -