Search

We introduce CaLMFlow (Causal Language Models for Flow Matching), a novel framework that casts flow matching as a Volterra integral equation (VIE), leveraging the power of large language models (LLMs) for continuous data generation. CaLMFlow enables the direct application of LLMs to learn complex flows by formulating flow matching as a sequence modeling task, bridging discrete language modeling and continuous generative modeling. Our method implements tokenization across space and time, thereby solving a VIE over these domains. This approach enables efficient handling of high-dimensional data and outperforms ODE solver-dependent methods like conditional flow matching (CFM). We demonstrate CaLMFlow's effectiveness on synthetic and real-world data, including single-cell perturbation response prediction, showcasing its ability to incorporate textual context and generalize to unseen conditions. Our results highlight LLM-driven flow matching as a promising paradigm in generative modeling, offering improved scalability, flexibility, and context-awareness., Comment: 10 pages, 9 figures, 7 tables

We explore the emergence of intelligent behavior in artificial systems by investigating how the complexity of rule-based systems influences the capabilities of models trained to predict these rules. Our study focuses on elementary cellular automata (ECA), simple yet powerful one-dimensional systems that generate behaviors ranging from trivial to highly complex. By training distinct Large Language Models (LLMs) on different ECAs, we evaluated the relationship between the complexity of the rules' behavior and the intelligence exhibited by the LLMs, as reflected in their performance on downstream tasks. Our findings reveal that rules with higher complexity lead to models exhibiting greater intelligence, as demonstrated by their performance on reasoning and chess move prediction tasks. Both uniform and periodic systems, and often also highly chaotic systems, resulted in poorer downstream performance, highlighting a sweet spot of complexity conducive to intelligence. We conjecture that intelligence arises from the ability to predict complexity and that creating intelligence may require only exposure to complexity., Comment: 15 pages,8 Figures

The significance of Mendeleev's periodic table extends beyond the classification of elements; it lies in its remarkable predictive power for discovering new elements and properties, revealing the underlying symmetrical patterns of nature that were only fully understood with the advent of quantum mechanics. Fundamental material properties, such as electron transport and magnetism, are also governed by crystal symmetry. In particular, spin transport depends on the spin polarization of electronic states, and recently discovered materials where the electron spin polarization is independent of momentum - a property known as a persistent spin texture (PST) - promise extended spin lifetime and efficient spin accumulation. In this paper, we establish the general conditions for the existence of symmetry-protected PST in bulk crystals. By systematically analyzing all 230 crystallographic space groups, similar to elements in the periodic table, we demonstrate that PST is universally present in all nonmagnetic solids lacking inversion symmetry. Using group theory, we identify the regions within the Brillouin zone that host PST and determine the corresponding directions of spin polarization. Our findings, supported by first-principles calculations of representative materials, open the route for discovering robust spintronic materials based on PST.

Data-driven technologies have improved the efficiency, reliability and effectiveness of healthcare services, but come with an increasing demand for data, which is challenging due to privacy-related constraints on sharing data in healthcare contexts. Synthetic data has recently gained popularity as potential solution, but in the flurry of current research it can be hard to oversee its potential. This paper proposes a novel taxonomy of synthetic data in healthcare to navigate the landscape in terms of three main varieties. Data Proportion comprises different ratios of synthetic data in a dataset and associated pros and cons. Data Modality refers to the different data formats amenable to synthesis and format-specific challenges. Data Transformation concerns improving specific aspects of a dataset like its utility or privacy with synthetic data. Our taxonomy aims to help researchers in the healthcare domain interested in synthetic data to grasp what types of datasets, data modalities, and transformations are possible with synthetic data, and where the challenges and overlaps between the varieties lie., Comment: Accepted at the 23rd EFMI Special Topic Conference, Romania, November 2024

Gold-standard sleep scoring as performed by human technicians is based on a subset of PSG signals, namely the EEG, EOG, and EMG. The PSG, however, consists of many more signal derivations that could potentially be used to perform sleep staging, including cardiac and respiratory modalities. Leveraging this variety in signals would offer advantages, for example by increasing reliability, resilience to signal loss, and application to long-term non-obtrusive recordings. This paper proposes a deep generative foundation model for fully automatic sleep staging from a plurality of sensors and any combination thereof. We trained a score-based diffusion model with a transformer backbone using a dataset of 1947 expert-labeled overnight sleep recordings with 36 different signals, including neurological, cardiac, and respiratory signals. We achieve zero-shot inference on any sensor set by using a novel Bayesian factorization of the score function across the sensors, i.e., it does not require retraining on specific combinations of signals. On single-channel EEG, our method reaches the performance limit in terms of PSG inter-rater agreement (5-class accuracy 85.6%, kappa 0.791). At the same time, the method offers full flexibility to use any sensor set derived from other modalities, for example, as typically used in home recordings that include finger PPG, nasal cannula and thoracic belt (5-class accuracy 79.0%, kappa of 0.697), or by combining derivations not typically used for sleep staging such as the tibialis and sternocleidomastoid EMG (5-class accuracy 71.0%, kappa of 0.575). Additionally, we propose a novel interpretability metric in terms of information gain per sensor and show that this is linearly correlated with classification performance. Lastly, our foundation model allows for post-hoc addition of entirely new sensor modalities by merely training a score estimator on the novel input.

Cherenkov threshold detectors (XCET) are used for identifying particles in the experimental areas at CERN. These detectors observe Cherenkov light emitted by charged particles travelling inside a pressurized gas vessel. A key component of the XCET detector is the 45-degree flat mirror reflecting the Cherenkov light towards the photomultiplier (PMT). A thorough analysis and optimization was conducted on the design and materials of this mirror, along with the surface coatings and coating techniques. A suitable manufacturing process was selected, and the first mirror prototype was produced, installed, and tested in the East Area at CERN. Experimental data obtained during beam tests is presented to assess the efficiency of the new coating and materials used., Comment: 15th International Particle Accelerator Conference,Nashville, TN

Natural language has the universal properties of being compositional and grounded in reality. The emergence of linguistic properties is often investigated through simulations of emergent communication in referential games. However, these experiments have yielded mixed results compared to similar experiments addressing linguistic properties of human language. Here we address representational alignment as a potential contributing factor to these results. Specifically, we assess the representational alignment between agent image representations and between agent representations and input images. Doing so, we confirm that the emergent language does not appear to encode human-like conceptual visual features, since agent image representations drift away from inputs whilst inter-agent alignment increases. We moreover identify a strong relationship between inter-agent alignment and topographic similarity, a common metric for compositionality, and address its consequences. To address these issues, we introduce an alignment penalty that prevents representational drift but interestingly does not improve performance on a compositional discrimination task. Together, our findings emphasise the key role representational alignment plays in simulations of language emergence., Comment: Appeared at the 13th edition of the Workshop on Cognitive Modeling and Computational Linguistics (CMCL 2024)

This paper presents initial results from the ESA-funded ``SUPPPPRESS'' project, which aims to develop high-performance liquid-crystal coronagraphs for direct imaging of Earth-like exoplanets in reflected light. The project focuses on addressing the significant challenge of polarization leakage in vector vortex coronagraphs (VVCs). We utilize newly manufactured multi-grating, liquid-crystal VVCs, consisting of a two- or three-element stack of vortex and grating patterns, to reduce this leakage to the $10^{-10}$ contrast level. We detail the experimental setups, including calibration techniques with polarization microscopes and Mueller matrix ellipsometers to enhance the direct-write accuracy of the liquid-crystal patterns. The performance testing of these coronagraph masks will be conducted on the THD2 high-contrast imaging testbed in Paris., Comment: 14 pages, 10 figures

General relativity is highly successful in explaining a wide range of gravitational phenomena including the gravitational waves emitted by binary systems and the shadows cast by supermassive black holes. From a modern perspective the theory is not fundamental though, but constitutes the lowest order term in an effective field theory description of the gravitational force. As a consequence, the gravitational dynamics should receive corrections by higher-derivative terms. This essay discusses structural aspects associated with these corrections and summarizes their imprint on static, spherically symmetric geometries. Along these lines, we critically reassess the common practice of using local field redefinitions in order to simplify the dynamics at the danger of shifting physics effects into sectors which are beyond the approximation under consideration., Comment: 10 pages. Essay received Honorable Mention at the Gravity Research Foundation 2024 Awards for Essays on Gravitation. v2: fixed typos and updated to match published version

Virtual experiments (VEs), a modern tool in metrology, can be used to help perform an uncertainty evaluation for the measurand. Current guidelines in metrology do not cover the many possibilities to incorporate VEs into an uncertainty evaluation, and it is often difficult to assess if the intended use of a VE complies with said guidelines. In recent work, it was shown that a VE can be used in conjunction with real measurement data and a Monte Carlo procedure to produce equal results to a supplement of the Guide to the Expression of Uncertainty in Measurement. However, this was shown only for linear measurement models. In this work, we extend this Monte Carlo approach to a common class of non-linear measurement models and more complex VEs, providing a reference approach for suitable uncertainty evaluations involving VEs. Numerical examples are given to show that the theoretical derivations hold in a practical scenario., Comment: Conference proceeding submitted to XXIV IMEKO World Congress 2024

In this work we challenge arguments for robot rights on metaphysical, ethical and legal grounds. Metaphysically, we argue that machines are not the kinds of things that may be denied or granted rights. Building on theories of phenomenology and post-Cartesian approaches to cognitive science, we ground our position in the lived reality of actual humans in an increasingly ubiquitously connected, controlled, digitized, and surveilled society. Ethically, we argue that, given machines current and potential harms to the most marginalized in society, limits on (rather than rights for) machines should be at the centre of current AI ethics debate. From a legal perspective, the best analogy to robot rights is not human rights but corporate rights, a highly controversial concept whose most important effect has been the undermining of worker, consumer, and voter rights by advancing the power of capital to exercise outsized influence on politics and law. The idea of robot rights, we conclude, acts as a smoke screen, allowing theorists and futurists to fantasize about benevolently sentient machines with unalterable needs and desires protected by law. While such fantasies have motivated fascinating fiction and art, once they influence legal theory and practice articulating the scope of rights claims, they threaten to immunize from legal accountability the current AI and robotics that is fuelling surveillance capitalism, accelerating environmental destruction, and entrenching injustice and human suffering., Comment: Published in First Monday special issue entitled "Ideologies of AI and the consolidation of power"

A search for hidden-charm pentaquark states decaying to a range of $\Sigma_{c}\bar{D}$ and $\Lambda_{c}\bar{D}$ final states, as well as doubly-charmed pentaquark states to $\Sigma_{c}D$ and $\Lambda_{c}^{+}D$, is made using samples of proton-proton collision data corresponding to an integrated luminosity of $5.7fb^{-1}$ recorded by the LHCb detector at $\sqrt{s} = 13Te\kern -0.1em V$. Since no significant signals are found, upper limits are set on the pentaquark yields relative to that of the $\Lambda_{c}^{+}$ baryon in the $\Lambda_{c}^{+}\to pK^{-}\pi^{+}$ decay mode. The known pentaquark states are also investigated, and their signal yields are found to be consistent with zero in all cases., Comment: All figures and tables, along with any supplementary material and additional information, are available at https://cern.ch/lhcbproject/Publications/p/LHCb-PAPER-2023-018.html (LHCb public pages)

Synthetic data from generative models emerges as the privacy-preserving data-sharing solution. Such a synthetic data set shall resemble the original data without revealing identifiable private information. The backbone technology of tabular synthesizers is rooted in image generative models, ranging from Generative Adversarial Networks (GANs) to recent diffusion models. Recent prior work sheds light on the utility-privacy tradeoff on tabular data, revealing and quantifying privacy risks on synthetic data. We first conduct an exhaustive empirical analysis, highlighting the utility-privacy tradeoff of five state-of-the-art tabular synthesizers, against eight privacy attacks, with a special focus on membership inference attacks. Motivated by the observation of high data quality but also high privacy risk in tabular diffusion, we propose DP-TLDM, Differentially Private Tabular Latent Diffusion Model, which is composed of an autoencoder network to encode the tabular data and a latent diffusion model to synthesize the latent tables. Following the emerging f-DP framework, we apply DP-SGD to train the auto-encoder in combination with batch clipping and use the separation value as the privacy metric to better capture the privacy gain from DP algorithms. Our empirical evaluation demonstrates that DP-TLDM is capable of achieving a meaningful theoretical privacy guarantee while also significantly enhancing the utility of synthetic data. Specifically, compared to other DP-protected tabular generative models, DP-TLDM improves the synthetic quality by an average of 35% in data resemblance, 15% in the utility for downstream tasks, and 50% in data discriminability, all while preserving a comparable level of privacy risk.

The influence of collisions on the growth rate of trapped-electron modes (TEMs) in core plasmas is assessed through both analytical linear gyrokinetics and linear gyrokinetic simulations. Both methods are applied to the magnetic geometry of the DIII-D tokamak, as well as the Helically Symmetric eXperiment (HSX) and Wendelstein 7-X (W7-X) stellarators, in the absence of temperature gradients. Here we analytically investigate the influence of collisions on the TEM eigenmode frequency by a perturbative approach in the response of trapped particles to the mode, using an energy-dependent Krook operator to model collisions. Although the resulting growth rates exceed perturbative thresholds, they reveal important qualitative dependencies: a geometry-dependent stabilization rate occurs for all wavenumbers at high collisionality, while at low collisionality, a geometry-sensitive mixture of collisionless, resonantly driven, and collisionally destabilized modes is found. Additionally, linear gyrokinetic simulations have been performed with a rigorous pitch-angle scattering operator for the same geometries. In the case of DIII-D and large wavenumber modes in HSX, the trends predicted by analytical theory are reproduced. Dissimilarities are, however, obtained in W7-X geometry and for low wavenumber modes in HSX, which are shown to be due to a collision-induced transition to the Universal Instability as the dominant instability at marginal collisionality., Comment: 24 pages, 46 figures

A robust authentication and authorization mechanism is imperative in modular system development, where modularity and modular thinking are pivotal. Traditional systems often employ identity modules responsible for authentication and token issuance. Tokens, representing user credentials, offer advantages such as reduced reliance on passwords, limited lifespan, and scoped access. Despite these benefits, the "bearer token" problem persists, leaving systems vulnerable to abuse if tokens are compromised. We propose a token-based authentication mechanism addressing modular systems' critical bearer token problem. The proposed mechanism includes a novel RAF (Recursive Augmented Fernet) token, a blacklist component, and a policy enforcer component. RAF tokens are one-time-use tokens, like tickets. They carry commands, and the receiver of an RAF token can issue new tokens using the received RAF token. The blacklist component guarantees an RAF token can not be approved more than once, and the policy enforcer checks the compatibility of commands carried by an RAF token. We introduce two variations of RAF tokens: User-tied RAF, offering simplicity and compatibility, and Fully-tied RAF, providing enhanced security through service-specific secret keys. We thoroughly discuss the security guarantees, technical definitions, and construction of RAF tokens backed by game-based proofs. We demonstrate a proof of concept in the context of OpenStack, involving modifications to Keystone and creating an RAFT library. The experimental results reveal minimal overhead in typical scenarios, establishing the practicality and effectiveness of RAF. Our experiments show that the RAF mechanism beats the idea of using short-life Fernet tokens while providing much better security.

Unreliable XOR Arbiter PUFs were broken by a machine learning attack, which targets the underlying Arbiter PUFs individually. However, reliability information from the PUF was required for this attack. We show that, for the first time, a perfectly reliable XOR Arbiter PUF, where no reliability information is accessible, can be efficiently attacked in the same divide-and-conquer manner. Our key insight is that the responses of correlated challenges also reveal their distance to the decision boundary. This leads to a chosen challenge attack on XOR Arbiter PUFs. The effectiveness of our attack is confirmed through PUF simulation and FPGA implementation.

Developing new material platforms for use in superconductor-semiconductor hybrid structures is desirable due to limitations caused by intrinsic microwave losses present in commonly used III/V material systems. With the recent reports that tantalum provides drastic improvements when implemented in superconducting circuit elements over traditional Nb and Al films, exploring Ta as an alternative superconductor in hybrid material systems seems necessary. Here, we present our study on the growth of Ta on semiconducting Ge (001) substrates grown via molecular beam epitaxy. We show that the Ta diffuses into the Ge matrix in a self-limiting nature resulting in extremely smooth and abrupt surfaces and interfaces that are ideal for future cQED device fabrication. The films have a nominal composition of TaGe$_{2}$ and form a native oxide of nominal composition Ta$_{2}$Ge$_{2}$O$_{5}$ that also forms a sharp interface with the underlying film. These films are superconducting with a $T_{C}\sim 1.8-2$K and $H_{C}^{\perp} \sim 1.88T$, $H_{C}^{\parallel} \sim 5.1T$.