Search

The development of a kilometer-scale E3SM Land Model (km-scale ELM) is an integral part of the E3SM project, which seeks to advance energy-related Earth system science research with state-of-the-art modeling and simulation capabilities on exascale computing systems. Through the utilization of high-fidelity data products, such as atmospheric forcing and soil properties, the km-scale ELM plays a critical role in accurately modeling geographical characteristics and extreme weather occurrences. The model is vital for enhancing our comprehension and prediction of climate patterns, as well as their effects on ecosystems and human activities. This study showcases the first set of full-capability, km-scale ELM simulations over various computational domains, including simulations encompassing 21.6 million land gridcells, reflecting approximately 21.5 million square kilometers of North America at a 1 km x 1 km resolution. We present the largest km-scale ELM simulation using up to 100,800 CPU cores across 2,400 nodes. This continental-scale simulation is 300 times larger than any previous studies, and the computational resources used are about 400 times larger than those used in prior efforts. Both strong and weak scaling tests have been conducted, revealing exceptional performance efficiency and resource utilization. The km-scale ELM uses the common E3SM modeling infrastructure and a general data toolkit known as KiloCraft. Consequently, it can be readily adapted for both fully-coupled E3SM simulations and data-driven simulations over specific areas, ranging from a single gridcell to the entire North America.

Internships have become a vital stepping stone for students transitioning from college to career. They offer learners the chance to explore career paths, build essential skills, and increase their chances of securing employment after graduation. However, the demand for internships far exceeds the available supply, leaving millions of students without the opportunities they need to succeed. "Expanding Internships: Harnessing Employer Insights to Boost Opportunity and Enhance Learning" explores this critical issue. The report reveals that in 2023, while 3.6 million students completed internships, another 4.6 million who sought similar experiences were unable to participate, often due to barriers related to race, socioeconomic status, and educational background. Through in-depth research and interviews with employers, the report identifies the key challenges companies face in offering internships, from operational constraints to market uncertainties. It also highlights the importance of quality in internship programs--students who complete high-quality internships gain a significant advantage in the job market. To address these challenges, BHEF offers four actionable recommendations aimed at increasing the supply of internships, improving their quality, and ensuring that all students have access to these critical opportunities.

This document shares valuable insights from peer mentorship programs at institutions that participated in APLU's Powered by Publics initiative and serves as a comprehensive resource for other institutions aiming to develop or enhance similar initiatives. By showcasing best practices, case examples, and practical recommendations, this report aims to empower universities to implement effective, evidence-based peer mentoring programs that significantly contribute to student success and equitable outcomes.

We describe updated scientific goals for the wide-field, millimeter-wave survey that will be produced by the Simons Observatory (SO). Significant upgrades to the 6-meter SO Large Aperture Telescope (LAT) are expected to be complete by 2028, and will include a doubled mapping speed with 30,000 new detectors and an automated data reduction pipeline. In addition, a new photovoltaic array will supply most of the observatory's power. The LAT survey will cover about 60% of the sky at a regular observing cadence, with five times the angular resolution and ten times the map depth of Planck. The science goals are to: (1) determine the physical conditions in the early universe and constrain the existence of new light particles; (2) measure the integrated distribution of mass, electron pressure, and electron momentum in the late-time universe, and, in combination with optical surveys, determine the neutrino mass and the effects of dark energy via tomographic measurements of the growth of structure at $z < 3$; (3) measure the distribution of electron density and pressure around galaxy groups and clusters, and calibrate the effects of energy input from galaxy formation on the surrounding environment; (4) produce a sample of more than 30,000 galaxy clusters, and more than 100,000 extragalactic millimeter sources, including regularly sampled AGN light-curves, to study these sources and their emission physics; (5) measure the polarized emission from magnetically aligned dust grains in our Galaxy, to study the properties of dust and the role of magnetic fields in star formation; (6) constrain asteroid regoliths, search for Trans-Neptunian Objects, and either detect or eliminate large portions of the phase space in the search for Planet 9; and (7) provide a powerful new window into the transient universe on time scales of minutes to years, concurrent with observations from Rubin of overlapping sky., Comment: 44 pages, 7 figures; abstract slightly abridged. Author contributions to this paper are available at https://simonsobservatory.org/wp-content/uploads/2025/02/Author-contribution-statement-20250228.pdf

Diffusion models may be formulated as a time-indexed sequence of energy-based models, where the score corresponds to the negative gradient of an energy function. As opposed to learning the score directly, an energy parameterization is attractive as the energy itself can be used to control generation via Monte Carlo samplers. Architectural constraints and training instability in energy parameterized models have so far yielded inferior performance compared to directly approximating the score or denoiser. We address these deficiencies by introducing a novel training regime for the energy function through distillation of pre-trained diffusion models, resembling a Helmholtz decomposition of the score vector field. We further showcase the synergies between energy and score by casting the diffusion sampling procedure as a Feynman Kac model where sampling is controlled using potentials from the learnt energy functions. The Feynman Kac model formalism enables composition and low temperature sampling through sequential Monte Carlo., Comment: Initial submission to openreview on October 3, 2024 (https://openreview.net/forum?id=6GyX0YRw8P); accepted to AISTATS 2025

We study the theoretical foundations of composition in diffusion models, with a particular focus on out-of-distribution extrapolation and length-generalization. Prior work has shown that composing distributions via linear score combination can achieve promising results, including length-generalization in some cases (Du et al., 2023; Liu et al., 2022). However, our theoretical understanding of how and why such compositions work remains incomplete. In fact, it is not even entirely clear what it means for composition to "work". This paper starts to address these fundamental gaps. We begin by precisely defining one possible desired result of composition, which we call projective composition. Then, we investigate: (1) when linear score combinations provably achieve projective composition, (2) whether reverse-diffusion sampling can generate the desired composition, and (3) the conditions under which composition fails. Finally, we connect our theoretical analysis to prior empirical observations where composition has either worked or failed, for reasons that were unclear at the time., Comment: 9 pages, 7 figures. The first two authors contributed equally

Antimicrobial photodynamic therapy (APDT) is a promising antibiotic-free strategy for broad-spectrum infection control in chronic wounds, minimising bacterial resistance risks. However, rapid photosensitiser diffusion, tissue staining, side toxicity, and short-lived antimicrobial effects present significant clinical limitations for integrating APDT into wound dressings. To address these challenges, we present the design of a bespoke polyvinyl alcohol (PVA) derivative conjugated with both phenothiazine and methacrylate functionalities, enabling staining-free antibacterial photodynamic effects, cellular tolerability and processability into various wound dressing formats, including films, textile fibres and nanoscale coatings. Tosylation of PVA is leveraged for the covalent coupling of toluidine blue, as confirmed by UV-Vis spectroscopy and the minimal release of TB in vitro. UV-induced network formation is exploited to accomplish cast films and nanoscale integrated wound dressing coatings. UV curing is also successfully coupled with an in-house wet spinning process to realise individual, water-insoluble fibres as the building blocks of fibrous wound dressings. A fluorometric assay supports the generation of reactive oxygen species when the UV-cured samples are exposed to work, but not UV, light, yielding a mean log10 reduction of up to 2.13 in S. aureus, and the complete eradication of P. aeruginosa. Direct and extract cytotoxicity tests with UV-cured films and fibres demonstrate the viability of L929 fibroblasts following 60-min light irradiation and 72-hour cell culture. The bespoke molecular architecture, broad processability and cellular tolerability of this PVA derivative are highly attractive aiming to integrate durable staining-free photodynamic capability in a wide range of healthcare technologies, from chronic wound dressings up to minimally invasive localised therapy., Comment: 12 figures. Accepted by European Polymer Journal

The Simons Observatory (SO) is a ground-based cosmic microwave background (CMB) survey experiment that currently consists of three 0.42m small-aperture telescopes (SATs) and one 6m large-aperture telescope (LAT), located at an elevation of 5200m in the Atacama Desert in Chile. At the LAT's focal plane, SO will install >62,000 transition-edge sensor detectors across 13 optics tubes (OTs) within the Large Aperture Telescope Receiver (LATR), the largest cryogenic camera ever built to observe the CMB. Here we report on the validation of the LATR in the laboratory and the subsequent dark testing and validation within the LAT. We show that the LATR meets cryogenic, optical, and detector specifications required for high-sensitivity measurements of the CMB. At the time of writing, the LATR is installed in the LAT with six OTs (corresponding to >31,000 detectors), and the LAT mirrors and remaining seven OTs are undergoing development.

There is a growing realization that neutrinos can be used as a diagnostic tool to better understand the inner workings of a nuclear weapon. Robust estimates demonstrate that an Inverse Beta Decay (IBD) neutrino scintillation detector built at the Nevada Test Site of 1000-ton active target mass at a standoff distance of 500 m would detect thousands of neutrino events per kTe of nuclear yield. This would provide less than 4% statistical error on measured neutrino rate and 5% error on neutrino energy. Extrapolating this to an error on the test device explosive yield requires knowledge from evaluated nuclear databases, non-equilibrium fission rates, and assumptions on internal neutron fluxes. Initial calculations demonstrate that prompt neutrino rates from a short pulse of Pu-239 fission is about a factor of two less than that from a steady state assumption. As well, there are significant energy spectral differences as a function of time after the pulse that needs to be considered. In the absence of nuclear weapons testing, many of the technical and theoretical challenges of a full nuclear test could be mitigated with a low cost smaller scale 20 ton fiducial mass IBD demonstration detector placed near a TRIGA pulsed reactor. The short duty cycle and repeatability of pulses would provide critical real environment testing and the measured neutrino rate as a function of time data would provide unique constraints on fission databases and equilibrium assumptions.

Background: The circadian rhythm aligns physiology and behaviour with the 24-hour light-dark cycle, and its disruption is linked to neurological disorders such as epilepsy. However, how to best quantify circadian disruption remains unclear, as it can manifest across various properties and timescales. A promising but under-explored approach is to assess the intra-individual variability in circadian rhythms over timescales of weeks to years. This is yet to be studied in epilepsy. Methods: We retrospectively used wearable smartwatch data (Fitbit) from 143 people with epilepsy (PWE) and 31 controls. For each participant, we extracted the circadian oscillation underlying their heart rate time series and analysed the intra-individual variability of three circadian properties: period, acrophase, and amplitude. Findings: We found increased intra-individual variability in period (77 min vs. 62 min, z=3.32, p<0.001) and acrophase (68 min vs. 54 min, z=2.97, p=0.003) for PWE compared to controls, but not in amplitude (1.98 bpm vs. 2.05 bpm, z=-0.66, p=0.51). For PWE, we did not find any correlations between seizure frequency and intra-individual variability in circadian properties, or any difference between weeks with and without seizures. Interpretation: This finding indicates that the circadian rhythm of heart rate is more variable for people with epilepsy and that this can be detected using a wearable device. However, we were unable to find any associations with seizure frequency or occurrence, suggesting intra-individual variability could be another manifestation of epilepsy aetiology. Future work should investigate the combined role of anti-seizure medications, demographics, co-morbidities, and health behaviours in driving the increased intra-individual variability of circadian properties in epilepsy.

We generalize ultraproducts and local-global limits of graphs to hypergraphs and other structures. We show that the local statistics of an ultraproduct of a sequence of hypergraphs are the ultralimits of the local statistics of the hypergraphs. Using some standard results from model theory, we conclude that the space of (equivalence classes of) pmp hypergraphs with the topology of local-global convergence is compact, and that any countable set of local statistics for a pmp hypergraph can be realized as the statistics of a set of labellings (rather than just approximated) in a local-global equivalent hypergraph. We give two applications. First, we characterize those structures where any solution to the corresponding CSP can be turned into a measurable solution. These turn out to be the width-1 structures. We can also use the limit machinery to extract from this theorem a purely finitary characterizations of width-1 structures involving asymptotic solutions. Second, we prove two measurable versions of the Frankl--R\"odl matching theorem using measurable nibble and differential equation arguments. The measurable proofs are much softer than the purely finitary results. And, we can recover the finitary theorems using the limit machinery.

During clinical treatment for epilepsy, the area of the brain thought to be responsible for pathological activity is identified. This identification is typically performed through visual assessment of EEG recordings; however, this is time consuming and prone to subjective inconsistency. Automated onset localisation algorithms provide objective identification of the onset location by highlighting changes in signal features associated with seizure onset. In this work we investigate how methodological differences in such algorithms can result in different onset locations being identified. We analysed ictal intracranial EEG (icEEG) recordings in 16 subjects (100 seizures) with drug-resistant epilepsy from the SWEZ-ETHZ public database. We identified a series of key methodological differences that must be considered when designing or selecting an onset localisation algorithm. These differences were demonstrated using three distinct algorithms that capture different, but complementary, seizure onset features: Imprint, Epileptogenicity Index, and Low Entropy Map. We assessed methodological differences (or Decision Points), and their impact on the identified onset locations. Our independent application of all three algorithms to the same ictal icEEG dataset revealed low agreement between them: 27-60% of onset channels showed minimal or no overlap. Therefore, we investigated the effect of three key differences: (i) how to define a baseline, (ii) whether low-frequency components are considered, and finally (iii) whether electrodecrement is considered. Changes at each Decision Point were found to substantially influence resultant onset channels (r>0.3). Our results demonstrate how seemingly small methodological changes can result in large differences in onset locations. We propose that key Decision Points must be considered when using or designing an onset localisation algorithm.

Diffusion and flow-matching models achieve remarkable generative performance but at the cost of many sampling steps, this slows inference and limits applicability to time-critical tasks. The ReFlow procedure can accelerate sampling by straightening generation trajectories. However, ReFlow is an iterative procedure, typically requiring training on simulated data, and results in reduced sample quality. To mitigate sample deterioration, we examine the design space of ReFlow and highlight potential pitfalls in prior heuristic practices. We then propose seven improvements for training dynamics, learning and inference, which are verified with thorough ablation studies on CIFAR10 $32 \times 32$, AFHQv2 $64 \times 64$, and FFHQ $64 \times 64$. Combining all our techniques, we achieve state-of-the-art FID scores (without / with guidance, resp.) for fast generation via neural ODEs: $2.23$ / $1.98$ on CIFAR10, $2.30$ / $1.91$ on AFHQv2, $2.84$ / $2.67$ on FFHQ, and $3.49$ / $1.74$ on ImageNet-64, all with merely $9$ neural function evaluations.

The increased adoption of diffusion models in text-to-image generation has triggered concerns on their reliability. Such models are now closely scrutinized under the lens of various metrics, notably calibration, fairness, or compute efficiency. We focus in this work on two issues that arise when deploying these models: a lack of diversity when prompting images, and a tendency to recreate images from the training set. To solve both problems, we propose a method that coaxes the sampled trajectories of pretrained diffusion models to land on images that fall outside of a reference set. We achieve this by adding repellency terms to the diffusion SDE throughout the generation trajectory, which are triggered whenever the path is expected to land too closely to an image in the shielded reference set. Our method is sparse in the sense that these repellency terms are zero and inactive most of the time, and even more so towards the end of the generation trajectory. Our method, named SPELL for sparse repellency, can be used either with a static reference set that contains protected images, or dynamically, by updating the set at each timestep with the expected images concurrently generated within a batch. We show that adding SPELL to popular diffusion models improves their diversity while impacting their FID only marginally, and performs comparatively better than other recent training-free diversity methods. We also demonstrate how SPELL can ensure a shielded generation away from a very large set of protected images by considering all 1.2M images from ImageNet as the protected set.

Denoising autoencoders for signal processing applications have been shown to experience significant difficulty in learning to reconstruct radio frequency communication signals, particularly in the large sample regime. In communication systems, this challenge is primarily due to the need to reconstruct the modulated data stream which is generally highly stochastic in nature. In this work, we take advantage of this limitation by using the denoising autoencoder to instead remove interfering radio frequency communication signals while reconstructing highly structured FMCW radar signals. More specifically, in this work we show that a CNN-layer only autoencoder architecture can be utilized to improve the accuracy of a radar altimeter's ranging estimate even in severe interference environments consisting of a multitude of interference signals. This is demonstrated through comprehensive performance analysis of an end-to-end FMCW radar altimeter simulation with and without the convolutional layer-only autoencoder. The proposed approach significantly improves interference mitigation in the presence of both narrow-band tone interference as well as wideband QPSK interference in terms of range RMS error, number of false altitude reports, and the peak-to-sidelobe ratio of the resulting range profile. FMCW radar signals of up to 40,000 IQ samples can be reliably reconstructed., Comment: To be presented at MILCOM 2024, Washington DC

Determining stress and strain in a component of a fusion power plant involves defining boundary conditions for the mechanical equilibrium equations, implying the availability of a full reactor model for defining those conditions. To address this fundamental challenge of reactor design, a finite element method (FEM) model for the Mega-Ampere Spherical Tokamak Upgrade (MAST-U) fusion tokamak, operating at the Culham Campus of UKAEA, has been developed and applied to assess mechanical deformations, strain, and stress in the full tokamak structure, taken as a proxy for a fusion power plant. The model, handling 127 million finite elements using about 800 processors in parallel, illustrates the level of fidelity of structural simulations of a complex nuclear device made possible by the modern supercomputing systems. The model predicts gravitational and atmospheric pressure-induced deformations in broad agreement with observations, and enables computing the spectrum of acoustic vibrations of a tokamak, arising from mechanical disturbances like an earthquake or a plasma disruption. We introduce the notion of the density of stress to characterise the distribution of stress in the entire solid body of the tokamak, and to predict the magnitude and locations of stress concentrations. The model enables defining computational requirements for simulating a whole operating fusion power plant, and provides a digital foundation for the assessment of reactor performance as well as for specifying the relevant materials testing programme.

We study jamming of an OFDM-modulated signal which employs forward error correction coding. We extend this to leverage reinforcement learning with a contextual bandit to jam a 5G-based system implementing some aspects of the 5G protocol. This model introduces unreliable reward feedback in the form of ACK/NACK observations to the jammer to understand the effect of how imperfect observations of errors can affect the jammer's ability to learn. We gain insights into the convergence time of the jammer and its ability to jam a victim 5G waveform, as well as insights into the vulnerabilities of wireless communications for reinforcement learning-based jamming., Comment: Accepted to IEEE MILCOM, Washington DC, 2024

We study how the rigidity transition in a triangular lattice changes as a function of anisotropy by preferentially filling bonds on the lattice in one direction. We discover that the onset of rigidity in anisotropic spring networks arises in at least two steps, reminiscent of the two-step melting transition in two dimensional crystals. In particular, our simulations demonstrate that the percolation of stress-supporting bonds happens at different critical volume fractions along different directions. By examining each independent component of the elasticity tensor, we determine universal exponents and develop universal scaling functions to analyze isotropic rigidity percolation as a multicritical point. We expect that these results will be important for elucidating the underlying mechanical phase transitions governing the properties of biological materials ranging from the cytoskeletons of cells to the extracellular networks of tissues such as tendon where the networks are often preferentially aligned., Comment: 13 pages, 17 figures

Plasma detachment involves interactions of the plasma with impurities and neutral particles, leading to significant losses of plasma power, momentum, and particles. Accurate mapping of plasma emissivity in the divertor and X-point region is essential for assessing the relationship between particle flux and radiative detachment. The recently validated InfraRed Video Bolometer (IRVB) diagnostic, in MAST-U enables this mapping with higher spatial resolution than more established methods like resistive bolometers. In previous preliminary work, the evolution of radiative detachment was characterised in L-mode (power entering the scrape-off layer, PSOL ~0.4MW). With a conventional divertor the inner leg consistently detached ahead of the outer leg, and radiative detachment preceded particle flux detachment. This work presents results also from the third MAST-U experimental campaign, fuelled from the low field side instead of the high field side, including Ohmic and beam heated L-mode shots (with a power exiting the core up to PSOL ~1-1.5MW). The radiation peak moves upstream from the target at lower upstream densities than the ion target flux roll-over (typically considered the detachment onset), while the inner leg detaches before the outer one. The movement of the radiation is in partial agreement with the expectations from the DLS model, predicting a sudden shift from the target to the X-point. The energy confinement is found to be related to detachment, but there seems to be some margin between the radiation on the inner leg reaching the X-point and confinement being affected, a beneficial characteristic if it could be extrapolated to future reactors. For increasing PSOL the particle flux roll over is almost unaffected, while radiative detachment occurs at higher density in both legs, but much higher on the outer, suggesting an uneven distribution of the power exiting the core., Comment: Submitted to Nuclear Materials and Energy

Anti-seizure medications (ASMs) are the mainstay of treatment for epilepsy, yet their effect on seizure spread is not fully understood. Higher ASM doses have been associated with shorter and less severe seizures. Our objective was to test if this effect was due to limiting seizure spread through early termination of otherwise unchanged seizures. We retrospectively examined intracranial EEG (iEEG) recordings in 15 subjects that underwent ASM tapering during pre-surgical monitoring. We estimated ASM plasma concentrations based on pharmaco-kinetic modelling. In each subject, we identified seizures that followed the same onset and initial spread patterns, but some seizures terminated early (truncated seizures), and other seizures continued to spread (continuing seizures). We compared ASM concentrations at the times of truncated seizures and continuing seizures. We found no substantial difference between ASM concentrations when truncated vs. continuing seizures occurred (Mean difference = 4%, sd = 29%, p=0.6). Our results indicate that ASM did not appear to halt established seizures in this cohort. Further research is needed to understand how ASM may modulate seizure duration and severity.

Aerosol interactions with clouds represent a significant uncertainty in our understanding of the Earth system. Deep convective clouds may respond to aerosol perturbations in several ways that have proven difficult to elucidate with observations. Here, we leverage the two busiest maritime shipping lanes in the world, which emit aerosol particles and their precursors into an otherwise relatively clean tropical marine boundary layer, to make headway on the influence of aerosol on deep convective clouds. The recent seven-fold change in allowable fuel sulfur by the International Maritime Organization allows us to test the sensitivity of the lightning to changes in ship plume aerosol size distributions. We find that, across a range of atmospheric thermodynamic conditions, the previously documented enhancement of lightning over the shipping lanes has fallen by over 40%. The enhancement is therefore at least partially aerosol-mediated, a conclusion that is supported by observations of droplet number at cloud base, which show a similar decline over the shipping lane. These results have fundamental implications for our understanding of aerosol-cloud interactions, suggesting that deep convective clouds are impacted by the aerosol number distribution in the remote marine environment.

We introduce Imagen 3, a latent diffusion model that generates high quality images from text prompts. We describe our quality and responsibility evaluations. Imagen 3 is preferred over other state-of-the-art (SOTA) models at the time of evaluation. In addition, we discuss issues around safety and representation, as well as methods we used to minimize the potential harm of our models.

The use of soft robotics for real-world underwater applications is limited, even more than in terrestrial applications, by the ability to accurately measure and control the deformation of the soft materials in real time without the need for feedback from an external sensor. Real-time underwater shape estimation would allow for accurate closed-loop control of soft propulsors, enabling high-performance swimming and manoeuvring. We propose and demonstrate a method for closed-loop underwater soft robotic foil control based on a flexible capacitive e-skin and machine learning which does not necessitate feedback from an external sensor. The underwater e-skin is applied to a highly flexible foil undergoing deformations from 2% to 9% of its camber by means of soft hydraulic actuators. Accurate set point regulation of the camber is successfully tracked during sinusoidal and triangle actuation routines with an amplitude of 5% peak-to-peak and 10-second period with a normalised RMS error of 0.11, and 2% peak-to-peak amplitude with a period of 5 seconds with a normalised RMS error of 0.03. The tail tip deflection can be measured across a 30 mm (0.15 chords) range. These results pave the way for using e-skin technology for underwater soft robotic closed-loop control applications., Comment: 6 pages, 10 figures

The glassy thermal conductivities observed in crystalline inorganic perovskites such as Cs$_3$Bi$_2$I$_6$Cl$_3$ is perplexing and lacking theoretical explanations. Here, we first experimentally measure such its thermal transport behavior from 20~K to 300~K, after synthesizing Cs$_3$Bi$_2$I$_6$Cl$_3$ single crystals. Using path-integral molecular dynamics simulations driven by machine learning potentials, we reveal that Cs$_3$Bi$_2$I$_6$Cl$_3$ has large lattice distortions at low temperatures, which may be related to the large atomic size mismatch. Employing the Wigner formulation of thermal transport, we reproduce the experimental thermal conductivities based on lattice-distorted structures. This study thus provides a framework for predicting and understanding glassy thermal transport in materials with strong lattice disorder.

Rigidity transitions induced by the formation of system-spanning disordered rigid clusters, like the jamming transition, can be well-described in most physically relevant dimensions by mean-field theories. A dynamical mean-field theory commonly used to study these transitions, the coherent potential approximation (CPA), shows logarithmic corrections in $2$ dimensions. By solving the theory in arbitrary dimensions and extracting the universal scaling predictions, we show that these logarithmic corrections are a symptom of an upper critical dimension $d_{u}=2$, below which the critical exponents are modified. We recapitulate Ken Wilson's phenomenology of the $(4-\epsilon)$-dimensional Ising model, but with the upper critical dimension reduced to $2$. We interpret this using normal form theory as a transcritical bifurcation in the RG flows and extract the universal nonlinear coefficients to make explicit predictions for the behavior near $2$ dimensions. This bifurcation is driven by a variable that is dangerously irrelevant in all dimensions $d>2$ which incorporates the physics of long-wavelength phonons and low-frequency elastic dissipation. We derive universal scaling functions from the CPA sufficient to predict all linear response in randomly diluted isotropic elastic systems in all dimensions., Comment: 22 pages, 6 figures. Revised abstract, added references

The Simons Observatory (SO) is a cosmic microwave background experiment composed of three 0.42 m Small Aperture Telescopes (SATs) and one 6 m Large Aperture Telescope (LAT) in the Atacama Desert of Chile. The Large Aperture Telescope Receiver (LATR) was integrated into the LAT in August 2023; however, because mirrors were not yet installed, the LATR optical chain was capped at the 4K stage. In this dark configuration we are able to characterize many elements of the instrument without contributions from atmospheric noise. Here we show this noise is below the required upper limit and its features are well described with a simple noise model. Maps produced using this noise model have properties that are in good agreement with the white noise levels of our dark data. Additionally, we show that our nominal scan strategy has a minimal effect on the noise when compared to the noise when the telescope is stationary

In this report, we introduce the Gemini 1.5 family of models, representing the next generation of highly compute-efficient multimodal models capable of recalling and reasoning over fine-grained information from millions of tokens of context, including multiple long documents and hours of video and audio. The family includes two new models: (1) an updated Gemini 1.5 Pro, which exceeds the February version on the great majority of capabilities and benchmarks; (2) Gemini 1.5 Flash, a more lightweight variant designed for efficiency with minimal regression in quality. Gemini 1.5 models achieve near-perfect recall on long-context retrieval tasks across modalities, improve the state-of-the-art in long-document QA, long-video QA and long-context ASR, and match or surpass Gemini 1.0 Ultra's state-of-the-art performance across a broad set of benchmarks. Studying the limits of Gemini 1.5's long-context ability, we find continued improvement in next-token prediction and near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 3.0 (200k) and GPT-4 Turbo (128k). Finally, we highlight real-world use cases, such as Gemini 1.5 collaborating with professionals on completing their tasks achieving 26 to 75% time savings across 10 different job categories, as well as surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person who learned from the same content.

Within the field of optimal transport (OT), the choice of ground cost is crucial to ensuring that the optimality of a transport map corresponds to usefulness in real-world applications. It is therefore desirable to use known information to tailor cost functions and hence learn OT maps which are adapted to the problem at hand. By considering a class of neural ground costs whose Monge maps have a known form, we construct a differentiable Monge map estimator which can be optimized to be consistent with known information about an OT map. In doing so, we simultaneously learn both an OT map estimator and a corresponding adapted cost function. Through suitable choices of loss function, our method provides a general approach for incorporating prior information about the Monge map itself when learning adapted OT maps and cost functions.

Optimal transport (OT) has profoundly impacted machine learning by providing theoretical and computational tools to realign datasets. In this context, given two large point clouds of sizes $n$ and $m$ in $\mathbb{R}^d$, entropic OT (EOT) solvers have emerged as the most reliable tool to either solve the Kantorovich problem and output a $n\times m$ coupling matrix, or to solve the Monge problem and learn a vector-valued push-forward map. While the robustness of EOT couplings/maps makes them a go-to choice in practical applications, EOT solvers remain difficult to tune because of a small but influential set of hyperparameters, notably the omnipresent entropic regularization strength $\varepsilon$. Setting $\varepsilon$ can be difficult, as it simultaneously impacts various performance metrics, such as compute speed, statistical performance, generalization, and bias. In this work, we propose a new class of EOT solvers (ProgOT), that can estimate both plans and transport maps. We take advantage of several opportunities to optimize the computation of EOT solutions by dividing mass displacement using a time discretization, borrowing inspiration from dynamic OT formulations, and conquering each of these steps using EOT with properly scheduled parameters. We provide experimental evidence demonstrating that ProgOT is a faster and more robust alternative to standard solvers when computing couplings at large scales, even outperforming neural network-based approaches. We also prove statistical consistency of our approach for estimating optimal transport maps., Comment: In proceedings of the 38th conference on Advances in Neural Information Processing Systems (NeurIPs), 23 pages, 14 figures

Learning meaningful representations of complex objects that can be seen through multiple ($k\geq 3$) views or modalities is a core task in machine learning. Existing methods use losses originally intended for paired views, and extend them to $k$ views, either by instantiating $\tfrac12k(k-1)$ loss-pairs, or by using reduced embeddings, following a \textit{one vs. average-of-rest} strategy. We propose the multi-marginal matching gap (M3G), a loss that borrows tools from multi-marginal optimal transport (MM-OT) theory to simultaneously incorporate all $k$ views. Given a batch of $n$ points, each seen as a $k$-tuple of views subsequently transformed into $k$ embeddings, our loss contrasts the cost of matching these $n$ ground-truth $k$-tuples with the MM-OT polymatching cost, which seeks $n$ optimally arranged $k$-tuples chosen within these $n\times k$ vectors. While the exponential complexity $O(n^k$) of the MM-OT problem may seem daunting, we show in experiments that a suitable generalization of the Sinkhorn algorithm for that problem can scale to, e.g., $k=3\sim 6$ views using mini-batches of size $64~\sim128$. Our experiments demonstrate improved performance over multiview extensions of pairwise losses, for both self-supervised and multimodal tasks., Comment: To be presented at ICML 2024

We report the results of an extensive set of simulations exploring the sensitivity of the BlackCAT CubeSat to long-duration gamma-ray bursts (GRBs). BlackCAT is a NASA APRA-funded CubeSat mission for the detection and real-time sub-arcminute localization of high-redshift ($z\gtrsim 3.5$) GRBs. Thanks to their luminous and long-lived afterglow emissions, GRBs are uniquely valuable probes of high-redshift star-forming galaxies and the intergalactic medium. In addition, each detected GRB with a known redshift serves to localize a region of high-redshift star formation in three dimensions, enabling deep follow-on searches for host galaxies and associated local and large-scale structures. We explore two distinct models for the GRB redshift distribution and luminosity function, both consistent with Swift observations. We find that, for either model, BlackCAT is expected to detect a mean of 42 bursts per year on-orbit, with 6.7% to 10% of these at $z>3.5$. BlackCAT bursts will be localized to $r_{90} \lesssim 55^{\prime\prime}$ precision and reported to the community within seconds. Due to the mission orbit and pointing scheme, bursts will be located in the night sky and well-placed for deep multiwavelength follow-up observations. BlackCAT is on schedule to achieve launch readiness in 2025., Comment: 10 pages, 4 figures, 3 tables; Published in ApJ

The Simons Observatory will map the temperature and polarization over half of the sky, at millimeter wavelengths in six spectral bands from the Atacama Desert in Chile. These data will provide new insights into the genesis, content, and history of our Universe; the astrophysics of galaxies and galaxy clusters; objects in our solar system; and time-varying astrophysical phenomena. This ambitious new instrument suite, initially comprising three 0.5 m small-aperture telescopes and one 6 m large aperture telescope, is designed using a common combination of new technologies and new implementations to realize an observatory significantly more capable than the previous generation. In this paper, we present the pre-deployment performance of the first mid-frequency "optics tube" which will be fielded on the large aperture telescope with sensitivity to the 90 and 150 GHz spectral bands. This optics tube contains lenses, filters, detectors, and readout components, all of which operate at cryogenic temperatures. It is one of seven that form the core of the large aperture telescope receiver in its initial deployment. We describe this optics tube, including details of comprehensive testing methods, new techniques for beam and passband characterization, and its measured performance. The performance metrics include beams, optical efficiency, passbands, and forecasts for the on-sky performance of the system. We forecast a sensitivity that exceeds the requirements of the large aperture telescope with greater than 30% margin in each spectral band, and predict that the instrument will realize diffraction-limited performance and the expected detector passbands.

Wearable sensors such as smartwatches have become ubiquitous in recent years, allowing the easy and continual measurement of physiological parameters such as heart rate, physical activity, body temperature, and blood glucose in an every-day setting. This multi-modal data offers the potential to identify latent states occurring across physiological measures, which may represent important bio-behavioural states that could not be observed in any single measure. Here we present an approach, utilising a hidden semi-Markov model, to identify such states in data collected using a smartwatch, electrocardiogram, and blood glucose monitor, over two weeks from a sample of 9 participants. We found 26 latent ultradian states across the sample, with many occurring at particular times of day. Here we describe some of these, as well as their association with subjective mood and time use diaries. These methods provide a novel avenue for developing insights into the physiology of everyday life.