Search

We provide a general theory of height-offset variables and their properties for nearest-neighbor integer-valued gradient models on trees. This notion goes back to Sheffield [25], who realized that such tail-measurable variables can be used to associate to gradient Gibbs measures also proper Gibbs measures, via the procedure of pinning at infinity. On the constructive side, our theory incorporates the existence of height-offset variables, regularity properties of their Lebesgue densities and concentration properties of the associated Gibbs measure. On the pathological side, we show that pinning at infinity necessarily comes at a cost. This phenomenon will be analyzed on the levels of translation invariance, the tree-indexed Markov chain property, and extremality. The scope of our theory incorporates free measures, and also height-periodic measures of period 2, assuming only finite second moments of the transfer operator which encodes the nearest neighbor interaction. Our proofs are based on investigations of the respective martingale limits, past and future tail-decompositions, and infinite product representations for moment generating functions., Comment: 27 pages, 2 figures

Field Programmable Gate Arrays (FPGAs) are known for their reprogrammability that allows for post-manufacture circuitry changes. Nowadays, they are integral to a variety of systems including high-security applications such as aerospace and military systems. However, this reprogrammability also introduces significant security challenges, as bitstream manipulation can directly alter hardware circuits. Malicious manipulations may lead to leakage of secret data and the implementation of hardware Trojans. In this paper, we present a comprehensive framework for manipulating bitstreams with minimal reverse engineering, thereby exposing the potential risks associated with inadequate bitstream protection. Our methodology does not require a complete understanding of proprietary bitstream formats or a fully reverse-engineered target design. Instead, it enables precise modifications by inserting pre-synthesized circuits into existing bitstreams. This novel approach is demonstrated through a semi-automated framework consisting of five steps: (1) partial bitstream reverse engineering, (2) designing the modification, (3) placing and (4) routing the modification into the existing circuit, and (5) merging of the modification with the original bitstream. We validate our framework through four practical case studies on the OpenTitan design synthesized for Xilinx 7-Series FPGAs. While current protections such as bitstream authentication and encryption often fall short, our work highlights and discusses the urgency of developing effective countermeasures. We recommend using FPGAs as trust anchors only when bitstream manipulation attacks can be reliably excluded.

Understanding mass, size, and surface mass density of giant molecular clouds (GMCs) in galaxies is key to insights into star formation processes. We analyze these in M33 using Herschel dust and archival IRAM 30m telescope data, compared to Milky Way CO data. A Dendrogram algorithm on a 2D dust map and a Xco factor map are used for M33 instead of a constant value. Dust and CO-derived values are similar, with mean radii of $\sim\,$$58\,$pc for the dust and $\sim\,$$68\,$pc for CO. Largest GMAs are about $150\,$pc in radius, similar to the Milky Way, suggesting a size-limiting process. M33 contains less massive, lower-density GMCs compared to the Milky Way. The highest mass GMCs observed in the Milky Way are mostly absent in M33. M33's mean surface mass density is much lower, due to the Milky Way's higher column densities despite similar GMC areas. No systematic gradients in M33's physical properties were found with galactocentric radius, but higher densities and masses near the center suggest increased star formation. In both galaxies, 30% of molecular mass is central. The GMC mass power-law spectrum index is $\alpha=2.3\pm0.1$ and $\alpha=1.9\pm0.1$ for dust and CO in M33, respectively. We conclude that M33 and Milky Way GMCs are mostly similar, though M33 lacks high-mass GMCs, with no clear explanation. GMC properties weakly depend on galactic environment, with stellar feedback as a factor needing further study.

We consider $\mathbb{Z}_q$-valued clock models on a regular tree, for general classes of ferromagnetic nearest neighbor interactions which have a discrete rotational symmetry. It has been proved recently that, at strong enough coupling, families of homogeneous Markov chain Gibbs states $\mu_A$ coexist whose single-site marginals concentrate on $A\subset \mathbb{Z}_q$, and which are not convex combinations of each other [AbHeKuMa24]. In this note, we aim at a description of the extremal decomposition of $\mu_A$ for $|A|\geq 2$ into all extremal Gibbs measures, which may be spatially inhomogeneous. First, we show that in regimes of very strong coupling, $\mu_A$ is not extremal. Moreover, $\mu_A$ possesses a single-site reconstruction property which holds for spin values sent from the origin to infinity, when these initial values are chosen from $A$. As our main result, we show that $\mu_A$ decomposes into uncountably many extremal inhomogeneous states. The proof is based on multi-site reconstruction which allows to derive concentration properties of branch overlaps. Our method is based on a new good site/bad site decomposition adapted to the $A$-localization property, together with a coarse graining argument in local state space., Comment: 32 pages, 3 figures

The increasing importance of supply chain security for digital devices -- from consumer electronics to critical infrastructure -- has created a high demand for skilled cybersecurity experts. These experts use Hardware Reverse Engineering (HRE) as a crucial technique to ensure trust in digital semiconductors. Recently, the US and EU have provided substantial funding to educate this cybersecurity-ready semiconductor workforce, but success depends on the widespread availability of academic training programs. In this paper, we investigate the current state of education in hardware security and HRE to identify efficient approaches for establishing effective HRE training programs. Through a systematic literature review, we uncover 13 relevant courses, including eight with accompanying academic publications. We identify common topics, threat models, key pedagogical features, and course evaluation methods. We find that most hardware security courses do not prioritize HRE, making HRE training scarce. While the predominant course structure of lectures paired with hands-on projects appears to be largely effective, we observe a lack of standardized evaluation methods and limited reliability of student self-assessment surveys. Our results suggest several possible improvements to HRE education and offer recommendations for developing new training courses. We advocate for the integration of HRE education into curriculum guidelines to meet the growing societal and industry demand for HRE experts., Comment: 7 pages, 1 figure, 2 tables. To be published in Proceedings of the 56th ACM Technical Symposium on Computer Science Education V. 1. For supplementary materials, see https://osf.io/dt8ne/

Artificial intelligence (AI)-based decision support systems hold promise for enhancing diagnostic accuracy and efficiency in computational pathology. However, human-AI collaboration can introduce and amplify cognitive biases, such as confirmation bias caused by false confirmation when erroneous human opinions are reinforced by inaccurate AI output. This bias may worsen when time pressure, ubiquitously present in routine pathology, strains practitioners' cognitive resources. We quantified confirmation bias triggered by AI-induced false confirmation and examined the role of time constraints in a web-based experiment, where trained pathology experts (n=28) estimated tumor cell percentages. Our results suggest that AI integration may fuel confirmation bias, evidenced by a statistically significant positive linear-mixed-effects model coefficient linking AI recommendations mirroring flawed human judgment and alignment with system advice. Conversely, time pressure appeared to weaken this relationship. These findings highlight potential risks of AI use in healthcare and aim to support the safe integration of clinical decision support systems.

We consider nonlinear dispersive equations of Schr\"odinger-type involving fractional powers $0

Designing adaptive documents that are visually appealing across various devices and for diverse viewers is a challenging task. This is due to the wide variety of devices and different viewer requirements and preferences. Alterations to a document's content, style, or layout often necessitate numerous adjustments, potentially leading to a complete layout redesign. We introduce FlexDoc, a framework for creating and consuming documents that seamlessly adapt to different devices, author, and viewer preferences and interactions. It eliminates the need for manually creating multiple document layouts, as FlexDoc enables authors to define desired document properties using templates and employs both discrete and continuous optimization in a novel comprehensive optimization process, which leverages automatic text summarization and image carving techniques to adapt both layout and content during consumption dynamically. Furthermore, we demonstrate FlexDoc in multiple real-world application scenarios, such as news readers and academic papers.

The formalism to calculate excited state properties from the $GW$-Bethe-Salpeter equation (BSE) method is introduced, providing convenient access to excited state absorption, excited state circular dichroism, and excited state optical rotation in the framework of the $GW$-BSE method. This is achieved using the second-order transition density, which can be obtained by solving a set of auxiliary equations similar to time-dependent density functional theory (TD-DFT). The proposed formulation therefore leads to no increase in the formal computational complexity when compared to the corresponding ground state properties. We further outline the calculation of fully relaxed spin-orbit coupling matrix elements within the $GW$-BSE method, allowing us to include perturbative corrections for spin-orbit coupling in aforementioned properties. These corrections are also extended to TD-DFT. Excited state absorption and perturbative spin-orbit coupling corrections within $GW$-BSE are evaluated for a selected set of molecular systems, yielding promising results.

Coherence properties are central to quantum systems and are at the heart of phenomena such as superconductivity. Here we study coherence properties of an ultracold Bose gas in a two-dimensional optical lattice across the thermal phase transition. To infer the phase coherence and phase fluctuation profile, we use direct matter-wave imaging of higher Talbot revivals as well as a new phase microscope based on a site-resolved mapping of phase fluctuations to density fluctuations during matter-wave imaging. We observe the algebraic decay of the phase correlations in the superfluid phase and a linear temperature increase of the exponent. These techniques will also allow studying coherence properties in strongly-correlated quantum systems with full spatial resolution., Comment: 16 pages, 8 figures, including supplementary material

Modulation conditioned on measurements on entangled photonic quantum states is a cornerstone technology of optical quantum information processing. Performing this task with low latency requires combining single-photon-level detectors with both electronic logic processing and optical modulation in close proximity. In the technologically relevant telecom wavelength band, detection of photonic quantum states is best performed with high-efficiency, low-noise, and high-speed detectors based on the photon-induced breakdown of superconductivity. Therefore, using these devices for feedforward requires mutual compatibility of all components under cryogenic conditions. Here, we demonstrate low-latency feedforward using a quasi-photon-number-resolved measurement on a quantum light source. Specifically, we use a multipixel superconducting nanowire single-photon detector, amplifier, logic, and an integrated electro-optic modulator in situ below 4K. We modulate the signal mode of a spontaneous parametric down-conversion source, conditional on a photon-number measurement of the idler mode, with a total latency of (23+/-3)ns. The photon-number discrimination actively manipulates the signal mode photon statistics, which is itself a central component in photonic quantum computing reliant on heralded single-photon sources. This represents an important benchmark for the fastest quantum photonic feedforward experiments comprising measurement, amplification, logic and modulation. This has direct applications in quantum computing, communication, and simulation protocols.

This paper explores the impact of variable pragmatic competence on communicative success through simulating language learning and conversing between speakers and listeners with different levels of reasoning abilities. Through studying this interaction, we hypothesize that matching levels of reasoning between communication partners would create a more beneficial environment for communicative success and language learning. Our research findings indicate that learning from more explicit, literal language is advantageous, irrespective of the learner's level of pragmatic competence. Furthermore, we find that integrating pragmatic reasoning during language learning, not just during evaluation, significantly enhances overall communication performance. This paper provides key insights into the importance of aligning reasoning levels and incorporating pragmatic reasoning in optimizing communicative interactions., Comment: EMNLP 2024 main

We demonstrate a high-dimensional mode-sorter for single photons based on a multi-output quantum pulse gate, which we can program to switch between different temporal-mode encodings including pulse modes, frequency bins, time bins, and their superpositions. This device can facilitate practical realizations of quantum information applications such as high-dimensional quantum key distribution, and thus enables secure communication with enhanced information capacity. We characterize the mode-sorter through a detector tomography in 3 and 5 dimensions and find a fidelity up to $0.958\pm 0.030$ at the single-photon level.

We develop a semismooth Newton framework for the numerical solution of fixed-point equations that are posed in Banach spaces. The framework is motivated by applications in the field of obstacle-type quasi-variational inequalities and implicit obstacle problems. It is discussed in a general functional analytic setting and allows for inexact function evaluations and Newton steps. Moreover, if a certain contraction assumption holds, we show that it is possible to globalize the algorithm by means of the Banach fixed-point theorem and to ensure $q$-superlinear convergence to the problem solution for arbitrary starting values. By means of a localization technique, our Newton method can also be used to determine solutions of fixed-point equations that are only locally contractive and not uniquely solvable. We apply our algorithm to a quasi-variational inequality which arises in thermoforming and which not only involves the obstacle problem as a source of nonsmoothness but also a semilinear PDE containing a nondifferentiable Nemytskii operator. Our analysis is accompanied by numerical experiments that illustrate the mesh-independence and $q$-superlinear convergence of the developed solution algorithm.

Extremely low-resource (XLR) languages lack substantial corpora for training NLP models, motivating the use of all available resources such as dictionaries and grammar books. Machine Translation from One Book (Tanzer et al., 2024) suggests prompting long-context LLMs with one grammar book enables English-Kalamang translation, an unseen XLR language - a noteworthy case of linguistic knowledge helping an NLP task. We investigate whether the book's grammatical explanations or its parallel examples are most effective for learning XLR translation, finding almost all improvement stems from the parallel examples. Further, we find similar results for Nepali, a seen low-resource language, and achieve performance comparable to an LLM with a grammar book by simply fine-tuning an encoder-decoder translation model. We then investigate where grammar books help by testing two linguistic tasks, grammaticality judgment and gloss prediction, and we explore what kind of grammatical knowledge helps by introducing a typological feature prompt that achieves leading results on these more relevant tasks. We thus emphasise the importance of task-appropriate data for XLR languages: parallel examples for translation, and grammatical data for linguistic tasks. As we find no evidence that long-context LLMs can make effective use of grammatical explanations for XLR translation, we suggest data collection for multilingual XLR tasks such as translation is best focused on parallel data over linguistic description.

The conversion of solar energy into chemical energy, stored in the form of hydrogen, bears enormous potential as a sustainable fuel for powering emerging technologies. Photoactive oxynitrides are promising materials for splitting water into molecular oxygen and hydrogen. However, one of the issues limiting widespread commercial use of oxynitrides is the degradation during operation. While recent studies have shown the loss of nitrogen, its relation to the reduced efficiency has not been directly and systematically addressed with experiments. In this study, we demonstrate the impact of the anionic stoichiometry of BaTaO$_x$N$_y$ on its electronic structure and functional properties. Through experimental ion scattering, electron microscopy, and photoelectron spectroscopy investigations, we determine the anionic composition ranging from the bulk towards the surface of BaTaO$_x$N$_y$ thin films. This further serves as input for band structure computations modeling the substitutional disorder of the anion sites. Combining our experimental and computational approaches, we reveal the depth-dependent elemental composition of oxynitride films, resulting in downward band bending and the loss of semiconducting character towards the surface. Extending beyond idealized systems, we demonstrate the relation between the electronic properties of real oxynitride photoanodes and their performance, providing guidelines for engineering highly efficient photoelectrodes and photocatalysts for clean hydrogen production., Comment: 19 pages, 5 figures

We introduce the nested $p$-center problem, which is a multi-period variant of the well-known $p$-center problem. The use of the nesting concept allows to obtain solutions, which are consistent over the considered time horizon, i.e., facilities which are opened in a given time period stay open for subsequent time periods. This is important in real-life applications, as closing (and potential later re-opening) of facilities between time periods can be undesirable. We consider two different versions of our problem, with the difference being the objective function. The first version considers the sum of the absolute regrets (of nesting) over all time periods, and the second version considers minimizing the maximum relative regret over the time periods. We present three mixed-integer programming formulations for the version with absolute regret objective and two formulations for the version with relative regret objective. For all the formulations, we present valid inequalities. Based on the formulations and the valid inequalities, we develop branch-and-bound/branch-and-cut solution algorithms. These algorithms include a preprocessing procedure that exploits the nesting property and also begins heuristics and primal heuristics. We conducted a computational study on instances from the literature for the $p$-center problem, which we adapted to our problems. We also analyse the effect of nesting on the solution cost and the number of open facilities.

Trapped-ion hardware based on the Magnetic Gradient Induced Coupling (MAGIC) scheme is emerging as a promising platform for quantum computing. Nevertheless, in this (as in any other) quantum-computing platform, many technical questions still have to be resolved before large-scale and error-tolerant applications are possible. In this work, we present a thorough discussion of the contribution of higher-order terms to the MAGIC setup, which can occur due to anharmonicities in the external potential of the ion crystal (e.g., through Coulomb repulsion) or through curvature of the applied magnetic field. These terms take the form of three-spin couplings as well as diverse terms that couple spins to phonons. We find that most of these are negligible in realistic situations, with only two contributions that need careful attention. First, there are parasitic longitudinal fields whose strength increases with chain length, but which can easily be compensated by a microwave detuning. Second, anharmonicities of the Coulomb interaction can lead to well-known two-to-one conversions of phonon excitations, which can be avoided if the phonons are ground-state cooled. Our detailed analysis constitutes an important contribution on the way of making magnetic-gradient trapped-ion quantum technology fit for large-scale applications, and it may inspire new ways to purposefully design interaction terms., Comment: 14 pages, 4 figures

Ethereum smart contracts are executable programs deployed on a blockchain. Once deployed, they cannot be updated due to their inherent immutability. Moreover, they often manage valuable assets that are worth millions of dollars, making them attractive targets for attackers. The introduction of vulnerabilities in programs due to the reuse of vulnerable code posted on Q&A websites such as Stack Overflow is not a new issue. However, little effort has been made to analyze the extent of this issue on deployed smart contracts. In this paper, we conduct a study on the impact of vulnerable code reuse from Q&A websites during the development of smart contracts and provide tools uniquely fit to detect vulnerable code patterns in complete and incomplete Smart Contract code. This paper proposes a pattern-based vulnerability detection tool that is able to analyze code snippets (i.e., incomplete code) as well as full smart contracts based on the concept of code property graphs. We also propose a methodology that leverages fuzzy hashing to quickly detect code clones of vulnerable snippets among deployed smart contracts. Our results show that our vulnerability search, as well as our code clone detection, are comparable to state-of-the-art while being applicable to code snippets. Our large-scale study on 18,660 code snippets reveals that 4,596 of them are vulnerable, out of which 616 can be found in 7,852 deployed smart contracts. These results highlight that the reuse of vulnerable code snippets is indeed an issue in currently deployed smart contracts.

Conformal field theories with central charge $c\le1$ on random surfaces have been extensively studied in the past. Here, this discussion is extended from their equilibrium distribution to their critical dynamics. This is motivated by the conjecture that these models describe the time evolution of certain social networks that are self-driven to a critical point. The time evolution of the surface area is shown to follow a Cox Ingersol Ross process. Planar surfaces shrink, while higher genus surfaces grow until the cosmological constant stops their growth. Two different equilibrium states are distinguished, dominated by (i) planar surfaces, and (ii) ``foamy'' surfaces, whose genus diverges. Time variations of the order parameter are analyzed and are found to have generalized hyperbolic distributions. In state (i), those have power law tails with a tail index close to 4. Analogies between the time evolution of the order parameter and a multifractal random walk are also pointed out., Comment: 27 pages, 5 figures, revised section 4 ("Dynamic Genus")

This paper studies moving 180-degree N\'eel walls in ferromagnetic thin films under the reduced model for the in-plane magnetization proposed by Capella, Melcher and Otto [5], in the case when a sufficiently weak external magnetic field is applied. It is shown that the linearization around the moving N\'eel wall's phase determines a spectral problem that is a relatively bounded perturbation of the linearization around the static N\'eel wall, which is the solution when the external magnetic field is set to zero and which is spectrally stable. Uniform resolvent-type estimates for the linearized operator around the static wall are established in order to prove the spectral stability of the moving wall upon application of perturbation theory for linear operators. The spectral analysis is the basis to prove, in turn, both the decaying properties of the generated semigroup and the nonlinear stability of the moving N\'eel wall under small perturbations, in the case of a sufficiently weak external magnetic field. The stability of the static N\'eel wall, which was established in a companion paper [4], plays a key role to obtain the main result., Comment: 40 pages, 4 figures

Parameter-efficient finetuning (PEFT) methods effectively adapt large language models (LLMs) to diverse downstream tasks, reducing storage and GPU memory demands. Despite these advantages, several applications pose new challenges to PEFT beyond mere parameter efficiency. One notable challenge involves the efficient deployment of LLMs equipped with multiple task- or user-specific adapters, particularly when different adapters are needed for distinct requests within the same batch. Another challenge is the interpretability of LLMs, which is crucial for understanding how LLMs function. Previous studies introduced various approaches to address different challenges. In this paper, we introduce a novel method, RoAd, which employs a straightforward 2D rotation to adapt LLMs and addresses all the above challenges: (1) RoAd is remarkably parameter-efficient, delivering optimal performance on GLUE, eight commonsense reasoning tasks and four arithmetic reasoning tasks with $<0.1\%$ trainable parameters; (2) RoAd facilitates the efficient serving of requests requiring different adapters within a batch, with an overhead comparable to element-wise multiplication instead of batch matrix multiplication; (3) RoAd enhances LLM's interpretability through integration within a framework of distributed interchange intervention, demonstrated via composition experiments., Comment: Accepted to NeurIPS 2024. Code: https://github.com/BaohaoLiao/road

We investigate the dynamical and physical structures of bright-rimmed clouds (BRCs) in a nearby HII region. We focused on carbon- and oxygen-bearing species that trace photon-dominated regions (PDRs) and warm molecular cloud surfaces in order to understand the effect of UV radiation from the exciting stars on the cloud structure. We mapped four regions around the most prominent BRCs at scales of 4--10 arcmin in the HII region IC 1396 in [CII] 158 micron with (up)GREAT on board SOFIA. IC 1396 is predominantly excited by an O6.5V star. Toward IC 1396A, we also observed [OI] 63 micron and 145 micron. We combined these observations with JCMT archive data, which provide the low-J transitions of CO, $^{13}$CO, and C$^{18}$O. All spectra are velocity-resolved. The line profiles show a variety of velocity structures, which we investigated in detail for all observed emission lines. We find no clear sign of photoevaporating flows in the [CII] spectra, although the uncertainty in the location of the BRCs along the line of sight makes this interpretation inconclusive. Our analysis of the [$^{13}$CII] emission in IC 1396A suggests that the [CII] is likely mostly optically thin. The heating efficiency, measured by the ([CII]+[OI] 63 micron)/far-infrared intensity ratio, is higher in the northern part of IC 1396A than in the southern part, which may indicate a difference in the dust properties of the two areas. The complex velocity structures identified in the BRCs of IC 1396, which is apparently a relatively simple HII region, highlight the importance of velocity-resolved data for disentangling different components along the line of sight and thus facilitating a detailed study of the dynamics of the cloud. We also demonstrate that the optically thin [$^{13}$CII] and [OI] 145 micron emission lines are essential for a conclusive interpretation of the [CII] 158 micron and [OI] 63 micron line profiles., Comment: Accepted for publication in A&A, 26 pages including Appendices

The latest applications in ultrafast quantum metrology require bright, broadband bi-photon sources with one of the photons in the mid-infrared and the other in the visible to near infrared. However, existing sources based on bulk crystals are limited in brightness due to the short interaction length and only allow for limited dispersion engineering. Here, we present an integrated PDC source based on a Ti:LiNbO$_3$ waveguide that generates broadband bi-photons with central wavelengths at $860\,\mathrm{nm}$ and $2800\,\mathrm{nm}$. Their spectral bandwidth exceeds $25\,\mathrm{THz}$ and is achieved by simultaneous matching of the group velocities and cancellation of group velocity dispersion for the signal and idler field. We provide an intuitive understanding of the process by studying our source's behaviour at different temperatures and pump wavelengths, which agrees well with simulations.

This paper introduces two multilingual systems, IKUN and IKUN-C, developed for the general machine translation task in WMT24. IKUN and IKUN-C represent an open system and a constrained system, respectively, built on Llama-3-8b and Mistral-7B-v0.3. Both systems are designed to handle all 11 language directions using a single model. According to automatic evaluation metrics, IKUN-C achieved 6 first-place and 3 second-place finishes among all constrained systems, while IKUN secured 1 first-place and 2 second-place finishes across both open and constrained systems. These encouraging results suggest that large language models (LLMs) are nearing the level of proficiency required for effective multilingual machine translation. The systems are based on a two-stage approach: first, continuous pre-training on monolingual data in 10 languages, followed by fine-tuning on high-quality parallel data for 11 language directions. The primary difference between IKUN and IKUN-C lies in their monolingual pre-training strategy. IKUN-C is pre-trained using constrained monolingual data, whereas IKUN leverages monolingual data from the OSCAR dataset. In the second phase, both systems are fine-tuned on parallel data sourced from NTREX, Flores, and WMT16-23 for all 11 language pairs., Comment: typo: 120K -> 12K vocabulary size

Trusted execution environments (TEEs) protect the integrity and confidentiality of running code and its associated data. Nevertheless, TEEs' integrity protection does not extend to the state saved on disk. Furthermore, modern cloud-native applications heavily rely on orchestration (e.g., through systems such as Kubernetes) and, thus, have their services frequently restarted. During restarts, attackers can revert the state of confidential services to a previous version that may aid their malicious intent. This paper presents CRISP, a rollback protection mechanism that uses an existing runtime for Intel SGX and transparently prevents rollback. Our approach can constrain the attack window to a fixed and short period or give developers the tools to avoid the vulnerability window altogether. Finally, experiments show that applying CRISP in a critical stateful cloud-native application may incur a resource increase but only a minor performance penalty.

We propose that physical time is based on counting the oscillations of wave functions. The discrete counting of the ticks of these clocks does not depend on the metric frame. It remains well defined for the beginning epochs of the universe. The photon clock counts the oscillations of electromagnetic plane waves in the cosmic reference frame. It can be gauged with clocks for other massless or massive particles. These equivalent clocks form a clock system which defines a common universal continuous physical time. For the photon clock this physical time coincides with conformal time. The universe is eternal towards the past for typical models of inflationary cosmology -- the photon clock ticks an infinite number of times. We compare the photon clock system to geodesic physical time, which is a frame invariant generalization of proper time., Comment: 6 pages

Density functional theory has become the workhorse of quantum physics, chemistry, and materials science. Within these fields, a broad range of applications needs to be covered. These applications range from solids to molecular systems, from organic to inorganic chemistry, or even from electrons to other fermions such as protons or muons. This is emphasized by the plethora of density functional approximations that have been developed for various cases. In this work, a new local hybrid exchange-correlation density functional is constructed from first principles, promoting generality and transferability. We show that constraint satisfaction can be achieved even for admixtures with full exact exchange, without sacrificing accuracy. The performance of the new functional for electronic structure theory is assessed for thermochemical properties, excitation energies, M\"ossbauer isomer shifts, NMR spin--spin coupling constants, NMR shieldings and shifts, magnetizabilities, as well as EPR hyperfine coupling constants. Here, the new density functional shows excellent performance throughout all tests and is numerically robust only requiring small grids for converged results. Additionally, the functional can be easily generalized to arbitrary fermions as shown for electron-proton correlation energies. Therefore, we outline that density functionals generated in this way are general purpose tools for quantum mechanical studies.

Predicting chaotic dynamical systems is critical in many scientific fields such as weather prediction, but challenging due to the characterizing sensitive dependence on initial conditions. Traditional modeling approaches require extensive domain knowledge, often leading to a shift towards data-driven methods using machine learning. However, existing research provides inconclusive results on which machine learning methods are best suited for predicting chaotic systems. In this paper, we compare different lightweight and heavyweight machine learning architectures using extensive existing databases, as well as a newly introduced one that allows for uncertainty quantification in the benchmark results. We perform hyperparameter tuning based on computational cost and introduce a novel error metric, the cumulative maximum error, which combines several desirable properties of traditional metrics, tailored for chaotic systems. Our results show that well-tuned simple methods, as well as untuned baseline methods, often outperform state-of-the-art deep learning models, but their performance can vary significantly with different experimental setups. These findings underscore the importance of matching prediction methods to data characteristics and available computational resources.

Ternary transition metal nitrides (TMNs) promise to significantly expand the material design space by opening new functionality and enhancing existing properties. However, most systems have only been investigated computationally and limited understanding of their stabilizing mechanisms restricts translation to experimental synthesis. To better elucidate key factors in designing ternary TMNs, we experimentally fabricate and analyze the physical properties of the ternary Zr-Pt-N system. Structural analysis and DFT modeling demonstrate that Pt substitutes nitrogen on the non-metallic sublattice, which destabilizes the rock-salt structure and forms a complex cubic phase. We also show insolubility of Pt in the Zr-Pt-N at 45 at % with the formation of a secondary Pt-rich phase. The measured reduced plasma frequency, decrease in resistivity, and decrease in hardness reflect a dominance of metallic behavior in bonding. Contrary to previous computational predictions, Zr-Pt-N films are shown to be metastable systems where even low Pt concentrations (1%) facilitate a solid reaction with the Si-substrate, that is inaccessible in ZrN films., Comment: 6 pages, 3 figures

We noisily observe solutions of an ordinary differential equation $\dot u = f(u)$ at given times, where $u$ lives in a $d$-dimensional state space. The model function $f$ is unknown and belongs to a H\"older-type smoothness class with parameter $\beta$. For the nonparametric problem of estimating $f$, we provide lower bounds on the error in two complementary model specifications: the snake model with few, long observed solutions and the stubble model with many short ones. The lower bounds are minimax optimal in some settings. They depend on various parameters, which in the optimal asymptotic regime leads to the same rate for the squared error in both models: it is characterized by the exponent $-2\beta/(2(\beta+1)+d)$ for the total number of observations $n$. To derive these results, we establish a master theorem for lower bounds in general nonparametric regression problems, which makes the proofs more comparable and seems to be a useful tool for future use.

We study nonparametric estimation in dynamical systems described by ordinary differential equations (ODEs). Specifically, we focus on estimating the unknown function $f \colon \mathbb{R}^d \to \mathbb{R}^d$ that governs the system dynamics through the ODE $\dot{u}(t) = f(u(t))$, where observations $Y_{j,i} = u_j(t_{j,i}) + \varepsilon_{j,i}$ of solutions $u_j$ of the ODE are made at times $t_{j,i}$ with independent noise $\varepsilon_{j,i}$. We introduce two novel models -- the Stubble model and the Snake model -- to mitigate the issue of observation location dependence on $f$, an inherent difficulty in nonparametric estimation of ODE systems. In the Stubble model, we observe many short solutions with initial conditions that adequately cover the domain of interest. Here, we study an estimator based on multivariate local polynomial regression and univariate polynomial interpolation. In the Snake model we observe few long trajectories that traverse the domain on interest. Here, we study an estimator that combines univariate local polynomial estimation with multivariate polynomial interpolation. For both models, we establish error bounds of order $n^{-\frac{\beta}{2(\beta +1)+d}}$ for $\beta$-smooth functions $f$ in an infinite dimensional function class of H\"older-type and establish minimax optimality for the Stubble model in general and for the Snake model under some conditions via comparison to lower bounds from parallel work.

This work proposes stochastic partial differential equations (SPDEs) as a practical tool to replicate clustering effects of more detailed particle-based dynamics. Inspired by membrane-mediated receptor dynamics on cell surfaces, we formulate a stochastic particle-based model for diffusion and pairwise interaction of particles, leading to intriguing clustering phenomena. Employing numerical simulation and cluster detection methods, we explore the approximation of the particle-based clustering dynamics through mean-field approaches. We find that SPDEs successfully reproduce spatiotemporal clustering dynamics, not only in the initial cluster formation period, but also on longer time scales where the successive merging of clusters cannot be tracked by deterministic mean-field models. The computational efficiency of the SPDE approach allows us to generate extensive statistical data for parameter estimation in a simpler model that uses a Markov jump process to capture the temporal evolution of the cluster number.

Quantum simulation is making a significant impact on scientific research. The prevailing tendency of the field is to build quantum simulators that get closer to real-world systems of interest, in particular electronic materials. However, progress in the microscopic design also provides an opportunity for an orthogonal research direction: building quantum many-body systems beyond real-world limitations. This colloquium takes this perspective: Concentrating on synthetic quantum matter in non-standard lattice geometries, such as fractal lattices or quasicrystals, higher-dimensional or curved spaces, it aims at providing a fresh introduction to the field of quantum simulation aligned with recent trends across various quantum simulation platforms, including atomic, photonic, and electronic devices. We also shine light on the novel phenomena which arise from these geometries: Condensed matter physicists may appreciate the variety of different localization properties as well as novel topological phases which are offered by such exotic quantum simulators. But also in the search of quantum models for gravity and cosmology, quantum simulators of curved spaces can provide a useful experimental tool., Comment: 32 pages, 11 figures

Using Stickelberger's theorem on Gauss sums, we show that if $F$ is a planar function on a finite field $\mathbb{F}_q$, then for all non-zero functions $G : \mathbb{F}_q \to \mathbb{F}_q$, we have \begin{equation*} \mathrm{deg } \ G \circ F - \mathrm{deg } \ G \le \frac{n(p-1)}{2}\,, \end{equation*} where $q = p^n$ with $p$ a prime and $n$ a positive integer, and $\mathrm{deg } \ F$ is the algebraic degree of $F$, i.e., the degree of the corresponding multivariate polynomial over $\mathbb{F}_p$. This bound leads to a simpler proof of the classification of planar polynomials over $\mathbb{F}_p$ and planar monomials over $\mathbb{F}_{p^2}$. As a new result, using the same degree bound, we complete the classification of planar monomials for all $n = 2^k$ with $p>5$ and $k$ a non-negative integer. Finally, we state a conjecture on the sum of the base-$p$ digits of integers modulo $q-1$ that implies the complete classification of planar monomials over finite fields of characteristic $p>5$.

If an ultraviolet fixed point renders quantum gravity renormalizable, the effective potential for a singlet scalar field -- the cosmon -- can be computed according to the corresponding scaling solution of the renormalization group equations. We associate the largest intrinsic mass scale generated by the flow away from the fixed point with the scale of present dark energy density or even smaller. This results in a highly predictive scenario for the evolution of dynamical dark energy. It solves the cosmological constant problem dynamically, and may be called "quantum gravity quintessence". A first setting without quantum scale symmetry violation in the neutrino sector could explain the present amount of dark energy, but fails for the constraints on its time evolution. In contrast, a logarithmic scale symmetry violation in the beyond standard model sector responsible for the neutrino masses induces a non-vanishing cosmon-neutrino coupling in the Einstein frame. This yields a cosmology similar to growing neutrino quintessence, which could be compatible with present observations. The small number of unknown parameters turns the scaling solution for quantum gravity into a fundamental explanation of dynamical dark energy which can be falsified., Comment: New parameters for figures which are closer to recent observation. Comparison with DESI results. 17 pages, 10 figures

The massive amounts of web-mined parallel data contain large amounts of noise. Semantic misalignment, as the primary source of the noise, poses a challenge for training machine translation systems. In this paper, we first study the impact of real-world hard-to-detect misalignment noise by proposing a process to simulate the realistic misalignment controlled by semantic similarity. After quantitatively analyzing the impact of simulated misalignment on machine translation, we show the limited effectiveness of widely used pre-filters to improve the translation performance, underscoring the necessity of more fine-grained ways to handle data noise. By observing the increasing reliability of the model's self-knowledge for distinguishing misaligned and clean data at the token-level, we propose a self-correction approach which leverages the model's prediction distribution to revise the training supervision from the ground-truth data over training time. Through comprehensive experiments, we show that our self-correction method not only improves translation performance in the presence of simulated misalignment noise but also proves effective for real-world noisy web-mined datasets across eight translation tasks.

Deep reinforcement learning (DRL) allows a system to interact with its environment and take actions by training an efficient policy that maximizes self-defined rewards. In autonomous driving, it can be used as a strategy for high-level decision making, whereas low-level algorithms such as the hybrid A* path planning have proven their ability to solve the local trajectory planning problem. In this work, we combine these two methods where the DRL makes high-level decisions such as lane change commands. After obtaining the lane change command, the hybrid A* planner is able to generate a collision-free trajectory to be executed by a model predictive controller (MPC). In addition, the DRL algorithm is able to keep the lane change command consistent within a chosen time-period. Traffic rules are implemented using linear temporal logic (LTL), which is then utilized as a reward function in DRL. Furthermore, we validate the proposed method on a real system to demonstrate its feasibility from simulation to implementation on real hardware.

In the last years, Infrared Reflection-Absorption Spectroscopy (IRRAS) became a standard technique to study vibrational excitations of molecules. These investigations are strongly motivated by perspective applications in monitoring chemical processes. For a better understanding of the adsorption mechanism of molecules on dielectrics, the polarization-dependence of an interaction of infrared light with adsorbates at dielectric surfaces is commonly used. Thus, the peak positions in absorption spectra could be different for s- and p-polarized light. This shift between the peak positions depends on both the molecule itself and the dielectric substrate. While the origin of this shift is well understood for infinite two-dimensional adsorbate layers, finite-size samples, which consist of 2D islands of a small number of molecules, have never been considered. Here, we present a study on polarization-dependent finite-size effects in the optical response of such islands on dielectric substrates. The study uses a multi-scale modeling approach that connects quantum chemistry calculations to Maxwell scattering simulations. We distinguish the optical response of a single molecule, a finite number of molecules, and a two-dimensional adsorbate layer. We analyze CO and CO$_2$ molecules deposited on CeO$_2$ and Al$_2$O$_3$ substrates. The evolution of the shift between the polarization-dependent absorbance peaks is firstly studied for a single molecule, which it does not exhibit for at all, and for finite molecular islands, which it increases with increasing island size for, as well as for an infinite two-dimensional adsorbate layer. In the latter case, the agreement between the obtained results and the experimental IRRAS data and more traditional three/four-layer-model theoretical studies supports the predictive power of the multi-scale approach.

Following multiple instructions is a crucial ability for large language models (LLMs). Evaluating this ability comes with significant challenges: (i) limited coherence between multiple instructions, (ii) positional bias where the order of instructions affects model performance, and (iii) a lack of objectively verifiable tasks. To address these issues, we introduce a benchmark designed to evaluate models' abilities to follow multiple instructions through sequential instruction following (SIFo) tasks. In SIFo, the successful completion of multiple instructions is verifiable by examining only the final instruction. Our benchmark evaluates instruction following using four tasks (text modification, question answering, mathematics, and security rules), each assessing different aspects of sequential instruction following. Our evaluation of popular LLMs, both closed-source and open-source, shows that more recent and larger models significantly outperform their older and smaller counterparts on the SIFo tasks, validating the benchmark's effectiveness. All models struggle with following sequences of instructions, hinting at an important lack of robustness of today's language models., Comment: EMNLP 2024 Findings

The count of mitotic figures (MFs) observed in hematoxylin and eosin (H&E)-stained slides is an important prognostic marker as it is a measure for tumor cell proliferation. However, the identification of MFs has a known low inter-rater agreement. Deep learning algorithms can standardize this task, but they require large amounts of annotated data for training and validation. Furthermore, label noise introduced during the annotation process may impede the algorithm's performance. Unlike H&E, the mitosis-specific antibody phospho-histone H3 (PHH3) specifically highlights MFs. Counting MFs on slides stained against PHH3 leads to higher agreement among raters and has therefore recently been used as a ground truth for the annotation of MFs in H&E. However, as PHH3 facilitates the recognition of cells indistinguishable from H&E stain alone, the use of this ground truth could potentially introduce noise into the H&E-related dataset, impacting model performance. This study analyzes the impact of PHH3-assisted MF annotation on inter-rater reliability and object level agreement through an extensive multi-rater experiment. We found that the annotators' object-level agreement increased when using PHH3-assisted labeling. Subsequently, MF detectors were evaluated on the resulting datasets to investigate the influence of PHH3-assisted labeling on the models' performance. Additionally, a novel dual-stain MF detector was developed to investigate the interpretation-shift of PHH3-assisted labels used in H&E, which clearly outperformed single-stain detectors. However, the PHH3-assisted labels did not have a positive effect on solely H&E-based models. The high performance of our dual-input detector reveals an information mismatch between the H&E and PHH3-stained images as the cause of this effect., Comment: 10 pages, 5 figures, 1 Table

The proliferation of drones, or unmanned aerial vehicles (UAVs), has raised significant safety concerns due to their potential misuse in activities such as espionage, smuggling, and infrastructure disruption. This paper addresses the critical need for effective drone detection and classification systems that operate independently of UAV cooperation. We evaluate various convolutional neural networks (CNNs) for their ability to detect and classify drones using spectrogram data derived from consecutive Fourier transforms of signal components. The focus is on model robustness in low signal-to-noise ratio (SNR) environments, which is critical for real-world applications. A comprehensive dataset is provided to support future model development. In addition, we demonstrate a low-cost drone detection system using a standard computer, software-defined radio (SDR) and antenna, validated through real-world field testing. On our development dataset, all models consistently achieved an average balanced classification accuracy of >= 85% at SNR > -12dB. In the field test, these models achieved an average balance accuracy of > 80%, depending on transmitter distance and antenna direction. Our contributions include: a publicly available dataset for model development, a comparative analysis of CNN for drone detection under low SNR conditions, and the deployment and field evaluation of a practical, low-cost detection system., Comment: 11 pages, 9 figures

Modeling structure and behavior of software systems plays a crucial role in the industrial practice of software engineering. As with other software engineering artifacts, software models are subject to evolution. Supporting modelers in evolving software models with recommendations for model completions is still an open problem, though. In this paper, we explore the potential of large language models for this task. In particular, we propose an approach, RAMC, leveraging large language models, model histories, and retrieval-augmented generation for model completion. Through experiments on three datasets, including an industrial application, one public open-source community dataset, and one controlled collection of simulated model repositories, we evaluate the potential of large language models for model completion with RAMC. We found that large language models are indeed a promising technology for supporting software model evolution (62.30% semantically correct completions on real-world industrial data and up to 86.19% type-correct completions). The general inference capabilities of large language models are particularly useful when dealing with concepts for which there are few, noisy, or no examples at all.

Neural manifolds are an attractive theoretical framework for characterizing the complex behaviors of neural populations. However, many of the tools for identifying these low-dimensional subspaces are correlational and provide limited insight into the underlying dynamics. The ability to precisely control this latent activity would allow researchers to investigate the structure and function of neural manifolds. Employing techniques from the field of optimal control, we simulate controlling the latent dynamics of a neural population using closed-loop, dynamically generated sensory inputs. Using a spiking neural network (SNN) as a model of a neural circuit, we find low-dimensional representations of both the network activity (the neural manifold) and a set of salient visual stimuli. With a data-driven latent dynamics model, we apply model predictive control (MPC) to provide anticipatory, optimal control over the trajectory of the circuit in a latent space. We are able to control the latent dynamics of the SNN to follow several reference trajectories despite observing only a subset of neurons and with a substantial amount of unknown noise injected into the network. These results provide a framework to experimentally test for causal relationships between manifold dynamics and other variables of interest such as organismal behavior and BCI performance.

While multilingual language models (MLMs) have been trained on 100+ languages, they are typically only evaluated across a handful of them due to a lack of available test data in most languages. This is particularly problematic when assessing MLM's potential for low-resource and unseen languages. In this paper, we present an analysis of existing evaluation frameworks in multilingual NLP, discuss their limitations, and propose several directions for more robust and reliable evaluation practices. Furthermore, we empirically study to what extent machine translation offers a {reliable alternative to human translation} for large-scale evaluation of MLMs across a wide set of languages. We use a SOTA translation model to translate test data from 4 tasks to 198 languages and use them to evaluate three MLMs. We show that while the selected subsets of high-resource test languages are generally sufficiently representative of a wider range of high-resource languages, we tend to overestimate MLMs' ability on low-resource languages. Finally, we show that simpler baselines can achieve relatively strong performance without having benefited from large-scale multilingual pretraining.

We study optimal control problems that are governed by semilinear elliptic partial differential equations that involve non-Lipschitzian nonlinearities. It is shown that, for a certain class of such PDEs, the solution map is Fr\'{e}chet differentiable even though the differential operator contains a nondifferentiable term. We exploit this effect to establish first-order necessary optimality conditions for minimizers of the considered control problems. The resulting KKT-conditions take the form of coupled PDE-systems that are posed in non-Muckenhoupt weighted Sobolev spaces and raise interesting questions regarding the regularity of optimal controls, the derivation of second-order optimality conditions, and the analysis of finite element discretizations.

This study is aimed to contribute to a more comprehensive understanding of the molecular hydrogen distribution in the galaxy M33 by introducing novel methods for generating high angular resolution (18.2$''$, equivalent to 75 pc) column density maps of molecular hydrogen ($N_{\rm H_2}$). M33 is a local group galaxy that has been observed with Herschel in the far-infrared wavelength range from 70 to 500 $\mu$m. Previous studies have presented total hydrogen column density maps ($N_{\rm H}$), using these FIR data (partly combined with mid-IR maps), employing various methods. We first performed a spectral energy distribution fit to the 160, 250, 350, and 500 $\mu$m continuum data obtain $N_{\rm H}$, using a technique similar to one previously reported in the literature. We also use a second method which involves translating only the 250 $\mu$m map into a $N_{\rm H}$ map at the same angular resolution. An $N_{\rm H_2}$ map via each method is then obtained by subtracting the HI component. Distinguishing our study from previous ones, we adopt a more versatile approach by considering a variable emissivity index, $\beta$ and dust absorption coefficient, $\kappa_0$. This choice enables us to construct a $\kappa_0$ map, thereby enhancing the depth and accuracy of our investigation of the hydrogen column density. We address the inherent biases and challenges within both methods (which give similar results) and compare them with existing maps available in the literature. Moreover, we calculate a map of the carbon monoxide CO-to-H$_2$ conversion factor ($X_\mathrm{CO}$ factor), which shows a strong dispersion around an average value of $1.8\times10^{20}\,\mathrm{cm^{-2}/(K\,km\,s^{-1})}$ throughout the disk. We obtain column density probability distribution functions (N-PDFs) from the $N_{\rm H}$, $N_{\rm H_2}$, and $N_{HI}$ maps and discuss their shape.

Fine-tuning large language models (LLMs) for machine translation has shown improvements in overall translation quality. However, it is unclear what is the impact of fine-tuning on desirable LLM behaviors that are not present in neural machine translation models, such as steerability, inherent document-level translation abilities, and the ability to produce less literal translations. We perform an extensive translation evaluation on the LLaMA and Falcon family of models with model size ranging from 7 billion up to 65 billion parameters. Our results show that while fine-tuning improves the general translation quality of LLMs, several abilities degrade. In particular, we observe a decline in the ability to perform formality steering, to produce technical translations through few-shot examples, and to perform document-level translation. On the other hand, we observe that the model produces less literal translations after fine-tuning on parallel data. We show that by including monolingual data as part of the fine-tuning data we can maintain the abilities while simultaneously enhancing overall translation quality. Our findings emphasize the need for fine-tuning strategies that preserve the benefits of LLMs for machine translation., Comment: Accepted to ACL 2024 (long, main). Latest version includes link to the IdiomsInCtx-MT dataset

A transducer is finite-valued if for some bound k, it maps any given input to at most k outputs. For classical, one-way transducers, it is known since the 80s that finite valuedness entails decidability of the equivalence problem. This decidability result is in contrast to the general case, which makes finite-valued transducers very attractive. For classical transducers, it is also known that finite valuedness is decidable and that any k-valued finite transducer can be decomposed as a union of k single-valued finite transducers. In this paper, we extend the above results to copyless streaming string transducers (SSTs), answering questions raised by Alur and Deshmukh in 2011. SSTs strictly extend the expressiveness of one-way transducers via additional variables that store partial outputs. We prove that any k-valued SST can be effectively decomposed as a union of k (single-valued) deterministic SSTs. As a corollary, we obtain equivalence of SSTs and two-way transducers in the finite-valued case (those two models are incomparable in general). Another corollary is an elementary upper bound for checking equivalence of finite-valued SSTs. The latter problem was already known to be decidable, but the proof complexity was unknown (it relied on Ehrenfeucht's conjecture). Finally, our main result is that finite valuedness of SSTs is decidable. The complexity is PSpace, and even PTime when the number of variables is fixed., Comment: This article is an extended version of the LICS'24 paper by the same name