@inproceedings{18066, abstract = {Graph games lie at the algorithmic core of many automated design problems in computer science. These are games usually played between two players on a given graph, where the players keep moving a token along the edges according to pre-determined rules (turn-based, concurrent, etc.), and the winner is decided based on the infinite path (aka play) traversed by the token from a given initial position. In bidding games, the players initially get some monetary budgets which they need to use to bid for the privilege of moving the token at each step. Each round of bidding affects the players' available budgets, which is the only form of update that the budgets experience. We introduce bidding games with charging where the players can additionally improve their budgets during the game by collecting vertex-dependent monetary rewards, aka the "charges." Unlike traditional bidding games (where all charges are zero), bidding games with charging allow non-trivial recurrent behaviors. For example, a reachability objective may require multiple detours to vertices with high charges to earn additional budget. We show that, nonetheless, the central property of traditional bidding games generalizes to bidding games with charging: For each vertex there exists a threshold ratio, which is the necessary and sufficient fraction of the total budget for winning the game from that vertex. While the thresholds of traditional bidding games correspond to unique fixed points of linear systems of equations, in games with charging, these fixed points are no longer unique. This significantly complicates the proof of existence and the algorithmic computation of thresholds for infinite-duration objectives. We also provide the lower complexity bounds for computing thresholds for Rabin and Streett objectives, which are the first known lower bounds in any form of bidding games (with or without charging), and we solve the following repair problem for safety and reachability games that have unsatisfiable objectives: Can we distribute a given amount of charge to the players in a way such that the objective can be satisfied?}, author = {Avni, Guy and Goharshady, Ehsan Kafshdar and Henzinger, Thomas A and Mallik, Kaushik}, booktitle = {35th International Conference on Concurrency Theory}, isbn = {9783959773393}, issn = {1868-8969}, location = {Calgary, Canada}, publisher = {Schloss Dagstuhl - Leibniz-Zentrum für Informatik}, title = {{Bidding games with charging}}, doi = {10.4230/LIPIcs.CONCUR.2024.8}, volume = {311}, year = {2024}, } @inproceedings{18067, abstract = {An automaton 𝒜 is history-deterministic if its nondeterminism can be resolved on the fly, only using the prefix of the word read so far. This mild form of nondeterminism has attracted particular attention for its applications in synthesis problems. An automaton 𝒜 is guidable with respect to a class C of automata if it can fairly simulate every automaton in C, whose language is contained in that of 𝒜. In other words, guidable automata are those for which inclusion and simulation coincide, making them particularly interesting for model-checking. We study the connection between these two notions, and specifically the question of when they coincide. For classes of automata on which they do, deciding guidability, an otherwise challenging decision problem, reduces to deciding history-determinism, a problem that is starting to be well-understood for many classes. We provide a selection of sufficient criteria for a class of automata to guarantee the coincidence of the notions, and use them to show that the notions coincide for the most common automata classes, among which are ω-regular automata and many infinite-state automata with safety and reachability acceptance conditions, including vector addition systems with states, one-counter nets, pushdown-, Parikh-, and timed-automata. We also demonstrate that history-determinism and guidability do not always coincide, for example, for the classes of timed automata with a fixed number of clocks.}, author = {Boker, Udi and Henzinger, Thomas A and Lehtinen, Karoliina and Prakash, Aditya}, booktitle = {35th International Conference on Concurrency Theory}, isbn = {9783959773393}, issn = {1868-8969}, location = {Calgary, Canada}, publisher = {Schloss Dagstuhl - Leibniz-Zentrum für Informatik}, title = {{History-determinism vs fair simulation}}, doi = {10.4230/LIPIcs.CONCUR.2024.12}, volume = {311}, year = {2024}, } @inproceedings{18068, abstract = {We study the following refinement relation between nondeterministic state-transition models: model ℬ strategically dominates model 𝒜 iff every deterministic refinement of 𝒜 is language contained in some deterministic refinement of ℬ. While language containment is trace inclusion, and the (fair) simulation preorder coincides with tree inclusion, strategic dominance falls strictly between the two and can be characterized as "strategy inclusion" between 𝒜 and ℬ: every strategy that resolves the nondeterminism of 𝒜 is dominated by a strategy that resolves the nondeterminism of ℬ. Strategic dominance can be checked in 2-ExpTime by a decidable first-order Presburger logic with quantification over words and strategies, called resolver logic. We give several other applications of resolver logic, including checking the co-safety, co-liveness, and history-determinism of boolean and quantitative automata, and checking the inclusion between hyperproperties that are specified by nondeterministic boolean and quantitative automata.}, author = {Henzinger, Thomas A and Mazzocchi, Nicolas Adrien and Sarac, Naci E}, booktitle = {35th International Conference on Concurrency Theory}, isbn = {9783959773393}, issn = {1868-8969}, location = {Calgary, Canada}, publisher = {Schloss Dagstuhl - Leibniz-Zentrum für Informatik}, title = {{Strategic dominance: A new preorder for nondeterministic processes}}, doi = {10.4230/LIPIcs.CONCUR.2024.29}, volume = {311}, year = {2024}, } @article{18069, abstract = {We present results from the JWST First Reionization Epoch Spectroscopically Complete Observations survey on the star-forming sequence (SFS) of galaxies at 1.0 < z < 1.7, around the peak of the cosmic star formation history. Star formation rates (SFRs) are measured from the redshifted, relatively dust-insensitive Paschen-α emission line, and stellar mass measurements include the F444W (4.4 μm; rest-frame H) band. We find SFRs of galaxies with log(M*/M⊙) > 9.5 that are lower than found in many earlier studies by up to 0.6 dex, but in good agreement with recent results obtained with the Prospector fitting framework. The difference (log(SFR(Paα)-SFR(Prospector)) is −0.09 ± 0.04 dex at 1010−11M⊙. We also measure the empirical relation between Paschen-α luminosity and rest-frame H-band magnitude and find that the scatter is only 0.04 dex lower than that of the SFR–M* relation and is much lower than the systematic differences among relations in the literature due to various methods of converting observed measurements to physical properties. We additionally identify examples of sources—that, with standard cutoffs via the UVJ diagram, would be deemed quiescent—with significant (log(sSFR)> −11 yr−1), typically extended, Paschen-α emission. Our results may be indicative of the potential unification of methods used to derive the SFS with careful selection of star-forming galaxies and independent SFR and stellar mass indicators.}, author = {Neufeld, Chloe and Van Dokkum, Pieter and Asali, Yasmeen and Covelo-Paz, Alba and Leja, Joel and Lin, Jamie and Matthee, Jorryt J and Oesch, Pascal A. and Reddy, Naveen A. and Shivaei, Irene and Whitaker, Katherine E. and Wuyts, Stijn and Brammer, Gabriel and Marchesini, Danilo and Maseda, Michael V. and Naidu, Rohan P. and Nelson, Erica J. and Velichko, Anna and Weibel, Andrea and Xiao, Mengyuan}, issn = {1538-4357}, journal = {Astrophysical Journal}, number = {2}, publisher = {IOP Publishing}, title = {{FRESCO: The Paschen-α star-forming sequence at cosmic noon}}, doi = {10.3847/1538-4357/ad6158}, volume = {972}, year = {2024}, } @inproceedings{18070, abstract = {Parallel SGD in a shared-memory setting is oft-represented by the popular Hogwild! algorithm, in which lock-free updates are asynchronously performed by multiple computing processes. Unfortunately, scaling Hogwild! to distributed workers is largely unexplored. Specifically, it is unknown if any adaptation of Hogwild! to the popular decentralized multi-GPU setting offers any competitive speedup, either empirically or theoretically. In this work, we investigate the potential of decentralizing Hogwild! by incorporating simultaneously (a) asynchronous local gradient updates on the shared memory of GPUs, and (b) non-blocking asynchronous decentralized federated averaging. A naive direct implementation shows degradation in performance, arising from scheduling overheads and concurrent write conflicts on GPUs. To mitigate these drawbacks, we investigate and propose a new method, based on careful block selection rules, which update only portions of the parameter vectors. Our experiments show that the resulting decentralized training method exhibits improved throughput and competitive accuracy for standard image classification benchmarks on the CIFAR-10, CIFAR-100, and Imagenet datasets. On the theoretical side, we prove that our method guarantees sublinear ergodic convergence rates for non-convex objectives.}, author = {Chatterjee, Bapi and Kungurtsev, Vyacheslav and Alistarh, Dan-Adrian}, booktitle = {Proceedings of the 44th International Conference on Distributed Computing Systems}, isbn = {9798350386059}, issn = {2575-8411}, location = {Jersey City, NJ, United States}, pages = {857--868}, publisher = {IEEE}, title = {{Federated SGD with local asynchrony}}, doi = {10.1109/ICDCS60910.2024.00084}, year = {2024}, } @inproceedings{18071, abstract = {Recent advancements on DAG-based consensus protocols allow for blockchains with improved metrics and properties, such as throughput and censorship-resistance. Variants of the Bullshark [18] consensus protocol are adopted for practical use by the Sui blockchain, for improved latency. However, the protocol is leader-based, and is strongly affected by crashed leaders that can lead to various performance issues, for example, decreased transaction throughput. In this paper, we propose HammerHead, a DAG-based consensus protocol, that is inspired by Carousel [8] and provides Leader-Utilization. Our proposal differs from Carousel, which is built for a chained consensus protocol; in HammerHead chain quality is inherited by the DAG. HammerHead needs to preserve safety and liveness, despite validators committing leader vertices asynchronously. The key idea is to update leader schedules dynamically, based on the validators' scores during the previous schedule. We implement HammerHead and show a minor improvement in performance for cases without faults. The major improvements in comparison to Bullshark appear in faulty settings. Specifically, we show a drastic, 2x-latency improvement and up to 40% increased throughput when crash faults occur (100 validators, 33 faults).}, author = {Tsimos, Giorgos and Kichidis, Anastasios and Sonnino, Alberto and Kokoris Kogias, Eleftherios}, booktitle = {Proceedings - International Conference on Distributed Computing Systems}, isbn = {9798350386059}, issn = {2575-8411}, location = {Jersey City, NJ, United States}, pages = {1377--1387}, publisher = {IEEE}, title = {{HammerHead: Leader reputation for dynamic scheduling}}, doi = {10.1109/ICDCS60910.2024.00129}, year = {2024}, } @article{18072, abstract = {The individualization of chromosomes during early mitosis and their clustering upon exit from cell division are two key transitions that ensure efficient segregation of eukaryotic chromosomes. Both processes are regulated by the surfactant-like protein Ki-67, but how Ki-67 achieves these diametric functions has remained unknown. Here, we report that Ki-67 radically switches from a chromosome repellent to a chromosome attractant during anaphase in human cells. We show that Ki-67 dephosphorylation during mitotic exit and the simultaneous exposure of a conserved basic patch induce the RNA-dependent formation of a liquid-like condensed phase on the chromosome surface. Experiments and coarse-grained simulations support a model in which the coalescence of chromosome surfaces, driven by co-condensation of Ki-67 and RNA, promotes clustering of chromosomes. Our study reveals how the switch of Ki-67 from a surfactant to a liquid-like condensed phase can generate mechanical forces during genome segregation that are required for re-establishing nuclear-cytoplasmic compartmentalization after mitosis.}, author = {Hernandez-Armendariz, Alberto and Sorichetti, Valerio and Hayashi, Yuki and Koskova, Zuzana and Brunner, Andreas and Ellenberg, Jan and Šarić, Anđela and Cuylen-Haering, Sara}, issn = {1097-4164}, journal = {Molecular Cell}, number = {17}, pages = {P3254--3270.E9}, publisher = {Cell Press}, title = {{A liquid-like coat mediates chromosome clustering during mitotic exit}}, doi = {10.1016/j.molcel.2024.07.022}, volume = {84}, year = {2024}, } @article{18073, abstract = {Conserved signaling cascades monitor protein-folding homeostasis to ensure proper cellular function. One of the evolutionary conserved key players is IRE1, which maintains endoplasmic reticulum (ER) homeostasis through the unfolded protein response (UPR). Upon accumulation of misfolded proteins in the ER, IRE1 forms clusters on the ER membrane to initiate UPR signaling. What regulates IRE1 cluster formation is not fully understood. Here, we show that the ER lumenal domain (LD) of human IRE1α forms biomolecular condensates in vitro. IRE1α LD condensates were stabilized both by binding to unfolded polypeptides as well as by tethering to model membranes, suggesting their role in assembling IRE1α into signaling-competent stable clusters. Molecular dynamics simulations indicated that weak multivalent interactions drive IRE1α LD clustering. Mutagenesis experiments identified disordered regions in IRE1α LD to control its clustering in vitro and in cells. Importantly, dysregulated clustering of IRE1α mutants led to defects in IRE1α signaling. Our results revealed that disordered regions in IRE1α LD control its clustering and suggest their role as a common strategy in regulating protein assembly on membranes.}, author = {Kettel, Paulina and Marosits, Laura and Spinetti, Elena and Rechberger, Michael and Giannini, Caterina and Radler, Philipp and Niedermoser, Isabell and Fischer, Irmgard and Versteeg, Gijs A. and Loose, Martin and Covino, Roberto and Karagöz, G. Elif}, issn = {1460-2075}, journal = {EMBO Journal}, number = {20}, pages = {4668--4698}, publisher = {Embo Press}, title = {{Disordered regions in the IRE1α ER lumenal domain mediate its stress-induced clustering}}, doi = {10.1038/s44318-024-00207-0}, volume = {43}, year = {2024}, } @phdthesis{18076, abstract = {The new era of Ge has opened up new possibilities in quantum computing. The maturity of Ge spin qubits is unquestioned, while hybrid semiconductor-superconductor Ge circuits are on track to enter the game. Gate-tunable transmons (gatemons) employing semiconductor Josephson junctions have recently emerged as building blocks for such hybrid quantum circuits. In this thesis, we present a gatemon fabricated in planar Germanium. We induce superconductivity in a two-dimensional hole gas by evaporating aluminum atop a thin spacer, which separates the superconductor from the Ge quantum well. The Josephson junction is then integrated into an Xmon circuit and capacitively coupled to a transmission line resonator. We showcase the qubit tunability in a broad frequency range with resonator and two-tone spectroscopy. Time-domain characterizations reveal energy relaxation and coherence times up to 75 ns. Our results, combined with the recent advances in the spin qubit field, pave the way towards novel hybrid and protected qubits in a group IV, CMOS-compatible material.}, author = {Sagi, Oliver}, issn = {2663-337X}, pages = {111}, publisher = {Institute of Science and Technology Austria}, title = {{Hybrid circuits on planar Germanium}}, doi = {10.15479/at:ista:18076}, year = {2024}, } @inproceedings{18086, abstract = {Abstract. Continuous group key agreement (CGKA) allows a group of users to maintain a continuously updated shared key in an asynchronous setting where parties only come online sporadically and their messages are relayed by an untrusted server. CGKA captures the basic primitive underlying group messaging schemes. Current solutions including TreeKEM (“Messaging Layer Security” (MLS) IETF RFC 9420) cannot handle concurrent requests while retaining low communication complexity. The exception being CoCoA, which is concurrent while having extremely low communication complexity (in groups of size n and for m concurrent updates the communication per user is log(n), i.e., independent of m). The main downside of CoCoA is that in groups of size n, users might have to do up to log(n) update requests to the server to ensure their (potentially corrupted) key material has been refreshed. In this work we present a “fast healing” concurrent CGKA protocol, named DeCAF, where users will heal after at most log(t) requests, with t being the number of corrupted users. While also suitable for the standard central-server setting, our protocol is particularly interesting for realizing decentralized group messaging, where protocol messages (add, remove, update) are being posted on some append-only data structure rather than sent to a server. In this setting, concurrency is crucial once the rate of requests exceeds, say, the rate at which new blocks are added to a blockchain. In the central-server setting, CoCoA (the only alternative with concurrency, sub-linear communication and basic post-compromise security) enjoys much lower download communication. However, in the decentralized setting – where there is no server which can craft specific messages for different users to reduce their download communication – our protocol significantly outperforms CoCoA. DeCAF heals in fewer epochs (log(t) vs. log(n)) while incurring a similar per epoch per user communication cost.}, author = {Alwen, Joel F and Auerbach, Benedikt and Cueto Noval, Miguel and Klein, Karen and Pascual Perez, Guillermo and Pietrzak, Krzysztof Z}, booktitle = {Security and Cryptography for Networks: 14th International Conference}, editor = {Galdi, Clemente and Phan, Duong Hieu}, isbn = {9783031710728}, issn = {1611-3349}, location = {Amalfi, Italy}, pages = {294–313}, publisher = {Springer Nature}, title = {{DeCAF: Decentralizable CGKA with fast healing}}, doi = {10.1007/978-3-031-71073-5_14}, volume = {14974}, year = {2024}, } @article{18087, abstract = {We present a theory describing the interaction of structured light, such as light carrying orbital angular momentum, with molecules. The light-matter interaction Hamiltonian we derive is expressed through couplings between spherical gradients of the electric field and the (transition) electric multipole moments of a particle of any nontrivial rotation point group. Our model can therefore accommodate an arbitrary complexity of the molecular and electric field structure, and it can be straightforwardly extended to atoms or nanostructures. Applying this framework to rovibrational spectroscopy of molecules, we uncover the general mechanism of angular momentum exchange between the spin and orbital angular momenta of light, molecular rotation, and its center-of-mass motion. We show that the nonzero vorticity of Laguerre-Gaussian beams can strongly enhance certain rovibrational transitions that are considered forbidden in the case of nonhelical light. We discuss the experimental requirements for the observation of these forbidden transitions in state-of-the-art spatially resolved spectroscopy measurements.}, author = {Maslov, Mikhail and Koutentakis, Georgios and Hrast, Mateja and Heckl, Oliver H. and Lemeshko, Mikhail}, issn = {2643-1564}, journal = {Physical Review Research}, number = {3}, publisher = {American Physical Society}, title = {{Theory of angular momentum transfer from light to molecules}}, doi = {10.1103/physrevresearch.6.033277}, volume = {6}, year = {2024}, } @phdthesis{18088, abstract = {Instant messaging applications like Whatsapp, Signal or Telegram have become ubiquitous in today's society. Many of them provide not only end-to-end encryption, but also security guarantees even when the key material gets compromised. These are achieved through frequent key update performed by users. In particular, the compromise of a group key should preserve confidentiality of previously exchanged messages (forward secrecy), and a subsequent key update will ensure security for future ones (post-compromise security). Though great protocols for one-on-one communication have been known for some time, the design of ones that scale efficiently for larger groups while achieving akin security guarantees is a hard problem. A great deal of research has been aimed at this topic, much of it under the umbrella of the Messaging Layer Security (MLS) working group at the IETF. Started in 2018, this joint effort by academics and industry culminated in 2023 with the publication of the first standard for secure group messaging [IETF, RFC9420]. At the core of secure group messaging is a cryptographic primitive termed Continuous Group Key Agreement, or CGKA [Alwen et al. 2021], that essentially allows a changing group of users to agree on a common key with the added functionality security against compromises is achieved by users asynchronously issuing a key update. In this thesis we contribute to the understanding of CGKA across different angles. First, we present a new technique to effect dynamic operations in groups, i.e., add or remove members, that can be more efficient that the one employed by MLS in certain settings. Considering the setting of users belonging to multiple overlapping groups, we then show lowerbounds on the communication cost of constructions that leverage said overlap, at the same time showing protocols that are asymptotically optimal and efficient for practical settings, respectively. Along the way, we show that the communication cost of key updates in MLS is average-cost optimal. An important feature in CGKA protocols, particularly for big groups, is the possibility of executing several group operations concurrently. While later versions of MLS support this, they do at the cost of worsening the communication efficiency of future group operations. In this thesis we introduce two new protocols that permit concurrency without any negative effect on efficiency. Our protocols circumvent previously existing lower bounds by satisfying a new notion of post-compromise security that only asks for security to be re-established after a certain number of key updates have taken place. While this can be slower than MLS in terms of rounds of communication, we show that it leads to more efficient overall communication. Additionally, we introduce a new technique that allows group members to decrease the information they need to store and download, which makes one of our protocols enjoy much lower download cost than any other existing CGKA constructions. }, author = {Pascual Perez, Guillermo}, issn = {2663-337X}, pages = {239}, publisher = {Institute of Science and Technology Austria}, title = {{On the efficiency and security of secure group messaging}}, doi = {10.15479/at:ista:18088}, year = {2024}, } @inproceedings{18097, abstract = {In our companion paper "Tight bounds for the learning of homotopy à la Niyogi, Smale, and Weinberger for subsets of Euclidean spaces and of Riemannian manifolds" we gave optimal bounds (in terms of the two one-sided Hausdorff distances) on a sample P of an input shape 𝒮 (either manifold or general set with positive reach) such that one can infer the homotopy of 𝒮 from the union of balls with some radius centred at P, both in Euclidean space and in a Riemannian manifold of bounded curvature. The construction showing the optimality of the bounds is not straightforward. The purpose of this video is to visualize and thus elucidate said construction in the Euclidean setting.}, author = {Attali, Dominique and Kourimska, Hana and Fillmore, Christopher D and Ghosh, Ishika and Lieutier, Andre and Stephenson, Elizabeth R and Wintraecken, Mathijs}, booktitle = {40th International Symposium on Computational Geometry}, isbn = {9783959773164}, location = {Athens, Greece}, publisher = {Schloss Dagstuhl - Leibniz-Zentrum für Informatik}, title = {{The ultimate frontier: An optimality construction for homotopy inference (media exposition)}}, doi = {10.4230/LIPIcs.SoCG.2024.87}, volume = {293}, year = {2024}, } @article{18107, abstract = {We consider a dilute fully spin-polarized Fermi gas at positive temperature in dimensions d∈{1,2,3} . We show that the pressure of the interacting gas is bounded from below by that of the free gas plus, to leading order, an explicit term of order adρ2+2/d, where a is the p-wave scattering length of the repulsive interaction and ρ is the particle density. The results are valid for a wide range of repulsive interactions, including that of a hard core, and uniform in temperatures at most of the order of the Fermi temperature. A central ingredient in the proof is a rigorous implementation of the fermionic cluster expansion of Gaudin, Gillespie and Ripka (Nucl. Phys. A, 176.2 (1971), pp. 237–260).}, author = {Lauritsen, Asbjørn Bækgaard and Seiringer, Robert}, issn = {2050-5094}, journal = {Forum of Mathematics, Sigma}, publisher = {Cambridge University Press}, title = {{Pressure of a dilute spin-polarized Fermi gas: Lower bound}}, doi = {10.1017/fms.2024.56}, volume = {12}, year = {2024}, } @article{18108, abstract = {Here we announce the construction and properties of a big commutative subalgebra of the Kirillov algebra attached to a finite dimensional irreducible representation of a complex semisimple Lie group. They are commutative finite flat algebras over the cohomology of the classifying space of the group. They are isomorphic with the equivariant intersection cohomology of affine Schubert varieties, endowing the latter with a new ring structure. Study of the finer aspects of the structure of the big algebras will also furnish the stalks of the intersection cohomology with ring structure, thus ringifying Lusztig’s q-weight multiplicity polynomials i.e., certain affine Kazhdan–Lusztig polynomials.}, author = {Hausel, Tamás}, issn = {1091-6490}, journal = {Proceedings of the National Academy of Sciences of the United States of America}, number = {38}, publisher = {National Academy of Sciences}, title = {{Commutative avatars of representations of semisimple Lie groups}}, doi = {10.1073/pnas.2319341121}, volume = {121}, year = {2024}, } @article{18109, abstract = {Venous thromboembolism (VTE) is a common, deadly disease with an increasing incidence despite preventive efforts. Clinical observations have associated elevated antibody concentrations or antibody-based therapies with thrombotic events. However, how antibodies contribute to thrombosis is unknown. Here, we show that reduced blood flow enabled immunoglobulin M (IgM) to bind to FcμR and the polymeric immunoglobulin receptor (pIgR), initiating endothelial activation and platelet recruitment. Subsequently, the procoagulant surface of activated platelets accommodated antigen- and FcγR-independent IgG deposition. This leads to classical complement activation, setting in motion a prothrombotic vicious circle. Key elements of this mechanism were present in humans in the setting of venous stasis as well as in the dysregulated immunothrombosis of COVID-19. This antibody-driven thrombosis can be prevented by pharmacologically targeting complement. Hence, our results uncover antibodies as previously unrecognized central regulators of thrombosis. These findings carry relevance for therapeutic application of antibodies and open innovative avenues to target thrombosis without compromising hemostasis.}, author = {Stark, Konstantin and Kilani, Badr and Stockhausen, Sven and Busse, Johanna and Schubert, Irene and Tran, Thuy Duong and Gärtner, Florian R and Leunig, Alexander and Pekayvaz, Kami and Nicolai, Leo and Fumagalli, Valeria and Stermann, Julia and Stephan, Felix and David, Christian and Müller, Martin B. and Heyman, Birgitta and Lux, Anja and Da Palma Guerreiro, Alexandra and Frenzel, Lukas P. and Schmidt, Christoph Q. and Dopler, Arthur and Moser, Markus and Chandraratne, Sue and Von Brühl, Marie Luise and Lorenz, Michael and Korff, Thomas and Rudelius, Martina and Popp, Oliver and Kirchner, Marieluise and Mertins, Philipp and Nimmerjahn, Falk and Iannacone, Matteo and Sperandio, Markus and Engelmann, Bernd and Verschoor, Admar and Massberg, Steffen}, issn = {1097-4180}, journal = {Immunity}, number = {9}, pages = {2140--2156}, publisher = {Elsevier}, title = {{Antibodies and complement are key drivers of thrombosis}}, doi = {10.1016/j.immuni.2024.08.007}, volume = {57}, year = {2024}, } @article{18110, abstract = {We study a chaotic particle-conserving kinetically constrained model, with a single parameter which allows us to break reflection symmetry. Through extensive numerical simulations we find that the domain wall state shows a variety of dynamical behaviors from localization all the way to ballistic transport, depending on the value of the reflection breaking parameter. Surprisingly, such anomalous behavior is not mirrored in infinite-temperature dynamics, which appear to scale diffusively, in line with expectations for generic interacting models. However, studying the particle density gradient, we show that the lack of reflection symmetry affects infinite-temperature dynamics, resulting in an asymmetric dynamical structure factor. This is in disagreement with normal diffusion and suggests that the model may also exhibit anomalous dynamics at infinite temperature in the thermodynamic limit. Finally, we observe low-entangled eigenstates in the spectrum of the model, a telltale sign of quantum many-body scars.}, author = {Brighi, Pietro and Ljubotina, Marko}, issn = {2469-9969}, journal = {Physical Review B}, number = {10}, publisher = {American Physical Society}, title = {{Anomalous transport in the kinetically constrained quantum East-West model}}, doi = {10.1103/PhysRevB.110.L100304}, volume = {110}, year = {2024}, } @article{18111, abstract = {Observations of tidal disruption events (TDEs) show signs of nitrogen enrichment reminiscent of other astrophysical sources such as active galactic nuclei and star-forming galaxies. Given that TDEs probe the gas from a single star, it is possible to test whether the observed enrichment is consistent with expectations from the CNO cycle by looking at the observed nitrogen/carbon (N/C) abundance ratios. Given that ≈20% of solar-mass stars (and an even larger fraction of more massive stars) live in close binaries, it is worthwhile to also consider what TDEs from stars influenced by binary evolution would look like. We show here that TDEs from stars stripped of their hydrogen-rich (and nitrogen-poor) envelopes through previous binary-induced mass loss can produce much higher observable N/C enhancements than even TDEs from massive stars. Additionally, we predict that the time dependence of the N/C abundance ratio in the mass fallback rate of stripped stars will follow the inverse behavior of main-sequence stars, enabling a more accurate characterization of the disrupted star.}, author = {Mockler, Brenna and Gallegos-Garcia, Monica and Götberg, Ylva Louise Linsdotter and Miller, Jon M. and Ramirez-Ruiz, Enrico}, issn = {2041-8213}, journal = {Astrophysical Journal Letters}, number = {1}, publisher = {IOP Publishing}, title = {{Tidal disruption events from stripped stars}}, doi = {10.3847/2041-8213/ad6c34}, volume = {973}, year = {2024}, } @inproceedings{18113, abstract = {The emergence of accurate open large language models (LLMs) has led to a race towards performant quantization techniques which can enable their execution on end-user devices. In this paper, we revisit the problem of “extreme” LLM compression—defined as targeting extremely low bit counts, such as 2 to 3 bits per parameter—from the point of view of classic methods in Multi-Codebook Quantization (MCQ). Our algorithm, called AQLM, generalizes the classic Additive Quantization (AQ) approach for information retrieval to advance the state-of-the-art in LLM compression, via two innovations: 1) learned additive quantization of weight matrices in input-adaptive fashion, and 2) joint optimization of codebook parameters across each transformer blocks. Broadly, AQLM is the first scheme that is Pareto optimal in terms of accuracy-vs-model-size when compressing to less than 3 bits per parameter, and significantly improves upon all known schemes in the extreme compression (2bit) regime. In addition, AQLM is practical: we provide fast GPU and CPU implementations of AQLM for token generation, which enable us to match or outperform optimized FP16 implementations for speed, while executing in a much smaller memory footprint.}, author = {Egiazarian, Vage and Panferov, Andrei and Kuznedelev, Denis and Frantar, Elias and Babenko, Artem and Alistarh, Dan-Adrian}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, issn = {2640-3498}, location = {Vienna, Austria}, pages = {12284--12303}, publisher = {ML Research Press}, title = {{Extreme compression of large language models via additive quantization}}, volume = {235}, year = {2024}, } @inproceedings{18114, abstract = {This paper presents Mechanistic Neural Networks, a neural network design for machine learning applications in the sciences. It incorporates a new Mechanistic Block in standard architectures to explicitly learn governing differential equations as representations, revealing the underlying dynamics of data and enhancing interpretability and efficiency in data modeling. Central to our approach is a novel Relaxed Linear Programming Solver (NeuRLP) inspired by a technique that reduces solving linear ODEs to solving linear programs. This integrates well with neural networks and surpasses the limitations of traditional ODE solvers enabling scalable GPU parallel processing. Overall, Mechanistic Neural Networks demonstrate their versatility for scientific machine learning applications, adeptly managing tasks from equation discovery to dynamic systems modeling. We prove their comprehensive capabilities in analyzing and interpreting complex scientific data across various applications, showing significant performance against specialized state-of-the-art methods. Source code is available at https://github.com/alpz/mech-nn.}, author = {Pervez, Adeel A and Locatello, Francesco and Gavves, Efstratios}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, issn = {2640-3498}, location = {Vienna, Austria}, pages = {40484--40501}, publisher = {ML Research Press}, title = {{Mechanistic neural networks for scientific machine learning}}, volume = {235}, year = {2024}, }