Search

With the growing attention and investment in recent AI approaches such as large language models, the narrative that the larger the AI system the more valuable, powerful and interesting it is is increasingly seen as common sense. But what is this assumption based on, and how are we measuring value, power, and performance? And what are the collateral consequences of this race to ever-increasing scale? Here, we scrutinize the current scaling trends and trade-offs across multiple axes and refute two common assumptions underlying the 'bigger-is-better' AI paradigm: 1) that improved performance is a product of increased scale, and 2) that all interesting problems addressed by AI require large-scale models. Rather, we argue that this approach is not only fragile scientifically, but comes with undesirable consequences. First, it is not sustainable, as its compute demands increase faster than model performance, leading to unreasonable economic requirements and a disproportionate environmental footprint. Second, it implies focusing on certain problems at the expense of others, leaving aside important applications, e.g. health, education, or the climate. Finally, it exacerbates a concentration of power, which centralizes decision-making in the hands of a few actors while threatening to disempower others in the context of shaping both AI research and its applications throughout society.

We evaluate the performance of various configurations of the Canadian Regional Climate Model (CRCM6-GEM5) in simulating 10-meter wind speeds using data from 27 AmeriFlux stations across North America. The assessment employs a hierarchy of error metrics, ranging from simple mean bias to advanced metrics that account for the dependence of wind speeds on variables such as friction velocity and stability. The results reveal that (i) the value of roughness length (z0) has a large effect on the simulation of wind speeds, (ii) using a lower limit for the Obhukov length instead of a lower limit for the lowest level wind speed seems to deteriorate the simulation of wind speeds under very stable conditions, (iii) the choice of stability function has a small but noticeable impact on the wind speeds, (iv) the turbulent orographic form drag scheme shows improvement over effective roughness length approach.

Causal structure learning (CSL) refers to the task of learning causal relationships from data. Advances in CSL now allow learning of causal graphs in diverse application domains, which has the potential to facilitate data-driven causal decision-making. Real-world CSL performance depends on a number of $\textit{context-specific}$ factors, including context-specific data distributions and non-linear dependencies, that are important in practical use-cases. However, our understanding of how to assess and select CSL methods in specific contexts remains limited. To address this gap, we present $\textit{CausalRegNet}$, a multiplicative effect structural causal model that allows for generating observational and interventional data incorporating context-specific properties, with a focus on the setting of gene perturbation experiments. Using real-world gene perturbation data, we show that CausalRegNet generates accurate distributions and scales far better than current simulation frameworks. We illustrate the use of CausalRegNet in assessing CSL methods in the context of interventional experiments in biology., Comment: 16 pages, 8 figures, 4 tables

PLATO (PLAnetary Transits and Oscillations of stars) is ESA's M3 mission designed to detect and characterise extrasolar planets and perform asteroseismic monitoring of a large number of stars. PLATO will detect small planets (down to <2 R_(Earth)) around bright stars (<11 mag), including terrestrial planets in the habitable zone of solar-like stars. With the complement of radial velocity observations from the ground, planets will be characterised for their radius, mass, and age with high accuracy (5 %, 10 %, 10 % for an Earth-Sun combination respectively). PLATO will provide us with a large-scale catalogue of well-characterised small planets up to intermediate orbital periods, relevant for a meaningful comparison to planet formation theories and to better understand planet evolution. It will make possible comparative exoplanetology to place our Solar System planets in a broader context. In parallel, PLATO will study (host) stars using asteroseismology, allowing us to determine the stellar properties with high accuracy, substantially enhancing our knowledge of stellar structure and evolution. The payload instrument consists of 26 cameras with 12cm aperture each. For at least four years, the mission will perform high-precision photometric measurements. Here we review the science objectives, present PLATO's target samples and fields, provide an overview of expected core science performance as well as a description of the instrument and the mission profile at the beginning of the serial production of the flight cameras. PLATO is scheduled for a launch date end 2026. This overview therefore provides a summary of the mission to the community in preparation of the upcoming operational phases.

We show that the dynamical system associated by Putnam to a pair of graph embeddings is identical to the shift map on the limit space of a self-similar groupoid action on a graph. Moreover, performing a certain out-split on said graph gives rise to a Katsura groupoid action on the out-split graph whose associated limit space dynamical system is conjugate to the previous one. We characterise the self-similar properties of these groupoids in terms of properties of their defining data, two matrices $A$, $B$. We prove a large class of the associated limit spaces are bundles of circles and points which fibre over a totally disconnected space, and the dynamics restricted to each circle is of the form $z\to z^{n}$. Moreover, we find a planar embedding of these spaces, thereby answering a question Putnam posed in his paper., Comment: 30 pages, 6 figures. Comments are welcome

We present the theory and experimental demonstration of a topological classification of finite tight binding Hamiltonians with chiral symmetry. Using the graph-theoretic notion of complete matchings, we show that many chiral tight binding structures can be divided into a number of sections, each of which has independent topological phases. Hence the overall classification is $N\mathbb{Z}_2$, corresponding to $2^N$ distinct phases, where $N$ is the number of sections with a non-trivial $\mathbb{Z}_2$ classification. In our classification, distinct topological phases are separated by exact closures in the energy spectrum of the Hamiltonian, with degenerate pairs of zero energy states. We show that that these zero energy states have an unusual localisation across distinct regions of the structure, determined by the manner in which the sections are connected together. We use this localisation to provide an experimental demonstration of the validity of the classification, through radio frequency measurements on a coaxial cable network which maps onto a tight binding system. The structure we investigate is a cable analogue of an ideal graphene ribbon, which divides into four sections and has a $4\mathbb{Z}_2$ topological classification., Comment: 18 pages, 9 figures, 2 tables

The miniaturisation of optical systems is an ongoing challenge across the electromagnetic spectrum. While the thickness of optical elements themselves can be reduced using advances in metamaterials, it is the voids between these elements -- which are necessary parts of an optical system -- that occupy most of the volume. Recently, a novel optical element coined a `spaceplate' was proposed, that replaces a region of free space with a thinner optical element that emulates the free-space optical response function -- thus having the potential to substantially shrink the volume of optical systems. While there have been a few proof-of-principle demonstrations of spaceplates, they have not yet been deployed in a real-world optical system. In this work, we use a bespoke-designed spaceplate to reduce the length of a gradient index (GRIN) lens microwave antenna. Our antenna is designed to operate at 23.5 GHz, and the incorporation of a nonlocal metamaterial spaceplate enables the distance between the antenna feed and the GRIN lens to be reduced by almost a factor of two. We find the radiation patterns from a conventional and space-squeezed antenna are very similar, with a very low cross-polarisation, and only a minor increase in the side-lobe levels when introducing the spaceplate. Our work represents a first example of a spaceplate integrated into a functional optical system, highlighting the potential for this concept to reduce the physical size of optical systems in real-world applications.

This paper explores the potential of communicating information gained by static analysis from compilers to Out-of-Order (OoO) machines, focusing on the memory dependence predictor (MDP). The MDP enables loads to issue without all in-flight store addresses being known, with minimal memory order violations. We use LLVM to find loads with no dependencies and label them via their opcode. These labelled loads skip making lookups into the MDP, improving prediction accuracy by reducing false dependencies. We communicate this information in a minimally intrusive way, i.e.~without introducing additional hardware costs or instruction bandwidth, providing these improvements without any additional overhead in the CPU. We find that in select cases in Spec2017, a significant number of load instructions can skip interacting with the MDP and lead to a performance gain. These results point to greater possibilities for static analysis as a source of near zero cost performance gains in future CPU designs., Comment: 15 pages

Language Models (LMs) such as BERT, have been shown to perform well on the task of identifying Named Entities (NE) in text. A BERT LM is typically used as a classifier to classify individual tokens in the input text, or to classify spans of tokens, as belonging to one of a set of possible NE categories. In this paper, we hypothesise that decoder-only Large Language Models (LLMs) can also be used generatively to extract both the NE, as well as potentially recover the correct surface form of the NE, where any spelling errors that were present in the input text get automatically corrected. We fine-tune two BERT LMs as baselines, as well as eight open-source LLMs, on the task of producing NEs from text that was obtained by applying Optical Character Recognition (OCR) to images of Japanese shop receipts; in this work, we do not attempt to find or evaluate the location of NEs in the text. We show that the best fine-tuned LLM performs as well as, or slightly better than, the best fine-tuned BERT LM, although the differences are not significant. However, the best LLM is also shown to correct OCR errors in some cases, as initially hypothesised., Comment: 9 pages, 1 figure

Quantum computers hold the promise of solving certain problems that lie beyond the reach of conventional computers. Establishing this capability, especially for impactful and meaningful problems, remains a central challenge. One such problem is the simulation of nonequilibrium dynamics of a magnetic spin system quenched through a quantum phase transition. State-of-the-art classical simulations demand resources that grow exponentially with system size. Here we show that superconducting quantum annealing processors can rapidly generate samples in close agreement with solutions of the Schr\"odinger equation. We demonstrate area-law scaling of entanglement in the model quench in two-, three- and infinite-dimensional spin glasses, supporting the observed stretched-exponential scaling of effort for classical approaches. We assess approximate methods based on tensor networks and neural networks and conclude that no known approach can achieve the same accuracy as the quantum annealer within a reasonable timeframe. Thus quantum annealers can answer questions of practical importance that classical computers cannot.

Colocalization analyses assess whether two traits are affected by the same or distinct causal genetic variants in a single gene region. A class of Bayesian colocalization tests are now routinely used in practice; for example, for genetic analyses in drug development pipelines. In this work, we consider an alternative frequentist approach to colocalization testing that examines the proportionality of genetic associations with each trait. The proportional colocalization approach uses markedly different assumptions to Bayesian colocalization tests, and therefore can provide valuable complementary evidence in cases where Bayesian colocalization results are inconclusive or sensitive to priors. We propose a novel conditional test of proportional colocalization, prop-coloc-cond, that aims to account for the uncertainty in variant selection, in order to recover accurate type I error control. The test can be implemented straightforwardly, requiring only summary data on genetic associations. Simulation evidence and an empirical investigation into GLP1R gene expression demonstrates how tests of proportional colocalization can offer important insights in conjunction with Bayesian colocalization tests.

We report an experimental study of the disordered Su-Schrieffer-Heeger (SSH) model, implemented in a system of coaxial cables, whose radio frequency properties map on to the SSH Hamiltonian. By measuring multiple chains with random hopping terms, we demonstrate the presence of a topologically protected state, with frequency variation of less than 0.2% over the ensemble. Connecting the ends of the chains to form loops, we observe a topological phase transition, characterised by the closure of the band gap and the appearance of states which are delocalised, despite the strong disorder.

Complex spatial and temporal structures are inherent characteristics of turbulent fluid flows and comprehending them poses a major challenge. This comprehesion necessitates an understanding of the space of turbulent fluid flow configurations. We employ a diffusion-based generative model to learn the distribution of turbulent vorticity profiles and generate snapshots of turbulent solutions to the incompressible Navier-Stokes equations. We consider the inverse cascade in two spatial dimensions and generate diverse turbulent solutions that differ from those in the training dataset. We analyze the statistical scaling properties of the new turbulent profiles, calculate their structure functions, energy power spectrum, velocity probability distribution function and moments of local energy dissipation. All the learnt scaling exponents are consistent with the expected Kolmogorov scaling. This agreement with established turbulence characteristics provides strong evidence of the model's capability to capture essential features of real-world turbulence.

Establishing lunar infrastructure is paramount to long-term habitation on the Moon. To meet the demand for future lunar infrastructure development, we present CraterGrader, a novel system for autonomous robotic earthmoving tasks within lunar constraints. In contrast to the current approaches to construction autonomy, CraterGrader uses online perception for dynamic mapping of deformable terrain, devises an energy-efficient material movement plan using an optimization-based transport planner, precisely localizes without GPS, and uses integrated drive and tool control to manipulate regolith with unknown and non-constant geotechnical parameters. We demonstrate CraterGrader's ability to achieve unprecedented performance in autonomous smoothing and grading within a lunar-like environment, showing that this framework is capable, robust, and a benchmark for future planetary site preparation robotics., Comment: 13 pages, 10 figures

Quantum Error Mitigation (QEM) presents a promising near-term approach to reduce error when estimating expectation values in quantum computing. Here, we introduce QEM techniques tailored for quantum annealing, using Zero-Noise Extrapolation (ZNE). We implement ZNE through zero-temperature extrapolation as well as energy-time rescaling. We conduct experimental investigations into the quantum critical dynamics of a transverse-field Ising spin chain, demonstrating the successful mitigation of thermal noise through both of these techniques. Moreover, we show that energy-time rescaling effectively mitigates control errors in the coherent regime where the effect of thermal noise is minimal. Our ZNE results agree with exact calculations of the coherent evolution over a range of annealing times that exceeds the coherent annealing range by almost an order of magnitude., Comment: 10 pages, 5 figures

A brief history of planar aperiodic tile sets is presented, starting from the Domino Problem proposed by Hao Wang in 1961. We provide highlights that led to the discovery of the Taylor--Socolar aperiodic monotile in 2010 and the Hat and Spectre aperiodic monotiles in 2023. The Spectre tile is an amazingly simple monotile; a single tile whose translated and rotated copies tile the plane but only in a way that lacks any translational periodicity. We showcase this breakthrough discovery through the 60$+$ years that aperiodic tile sets have been considered., Comment: Expository

There is wide agreement on the importance of implementation data from randomized effectiveness studies in behavioral science; however, there are few methods available to incorporate these data into causal models, especially when they are multivariate or longitudinal, and interest is in low-dimensional summaries. We introduce a framework for studying how treatment effects vary between subjects who implement an intervention differently, combining principal stratification with latent variable measurement models; since principal strata are latent in both treatment arms, we call it "fully-latent principal stratification" or FLPS. We describe FLPS models including item-response-theory measurement, show that they are feasible in a simulation study, and illustrate them in an analysis of hint usage from a randomized study of computerized mathematics tutors., Comment: In Submission

We report the observation of band gaps for low loss exciton-polaritons propagating outside the light cone in GaAs-based planar waveguides patterned into two-dimensional photonic crystals. By etching square lattice arrays of shallow holes into the uppermost layer of our structure, we open gaps on the order of 10 meV in the photonic mode dispersion, whose size and light-matter composition can be tuned by proximity to the strongly coupled exciton resonance. We demonstrate gaps ranging from almost fully photonic to highly excitonic. Opening a gap in the exciton-dominated part of the polariton spectrum is a promising first step towards the realization of quantum-Hall-like states arising from topologically nontrivial hybridization of excitons and photons., Comment: published in Appl. Phys. Lett. in 2021

Simultaneous localization and mapping (SLAM) is a fundamental task for numerous applications such as autonomous navigation and exploration. Despite many SLAM datasets have been released, current SLAM solutions still struggle to have sustained and resilient performance. One major issue is the absence of high-quality datasets including diverse all-weather conditions and a reliable metric for assessing robustness. This limitation significantly restricts the scalability and generalizability of SLAM technologies, impacting their development, validation, and deployment. To address this problem, we present SubT-MRS, an extremely challenging real-world dataset designed to push SLAM towards all-weather environments to pursue the most robust SLAM performance. It contains multi-degraded environments including over 30 diverse scenes such as structureless corridors, varying lighting conditions, and perceptual obscurants like smoke and dust; multimodal sensors such as LiDAR, fisheye camera, IMU, and thermal camera; and multiple locomotions like aerial, legged, and wheeled robots. We develop accuracy and robustness evaluation tracks for SLAM and introduced novel robustness metrics. Comprehensive studies are performed, revealing new observations, challenges, and opportunities for future research.

The exciton binding energy ($E_b$) is a key parameter that governs the physics of many optoelectronic devices. At their best, trustworthy and precise measurements of $E_b$ challenge theoreticians to refine models, are a driving force in advancing the understanding a material system, and lead to efficient device design. At their worst, inaccurate $E_b$ measurements lead theoreticians astray, sew confusion within the research community, and hinder device improvements by leading to poor designs. This review article seeks to highlight the pros and cons of different measurement techniques used to determine $E_b$, namely, temperature-dependent photoluminescence, resolving Rydberg states, electroabsorption, magnetoabsorption, scanning tunneling spectroscopy, and fitting the optical absorption. Due to numerous conflicting $E_b$ values reported for halide perovskites (HP) and transition metal dichalcogenides (TMDC) monolayers, an emphasis is placed on highlighting these measurements in attempt to reconcile the variance between different measurement techniques. By considering the published data en masse, we argue the experiments with the clearest indicators are in agreement on the following values: ~350 - 450 meV for TMDC monolayers between SiO$_2$ and vacuum, ~150 - 200 meV for hBN-encapsulated TMDC monolayers, ~200 - 300 meV for common lead-iodide 2D HPs, and ~10 meV for methylammonium lead iodide., Comment: 80 pages, 24 figures

Recently, the first JWST measurement of thermal emission from a rocky exoplanet was reported. The inferred dayside brightness temperature of TRAPPIST-1 b at 15 $\mu$m is consistent with the planet having no atmosphere and therefore no mechanism by which to circulate heat to its nightside. In this Letter, we compare the measured secondary eclipse depth of TRAPPIST-1 b to predictions from a suite of self-consistent radiative-convective equilibrium models in order to quantify the maximum atmospheric thickness consistent with the observation. We find that plausible atmospheres (i.e., those that contain at least 100 ppm CO$_2$) with surface pressures greater than 0.01 bar (0.1 bar) are ruled out at 1$\sigma$ (3$\sigma$), regardless of the choice of background atmosphere. Thicker atmospheres of up to 10 bar (100 bar) at 1$\sigma$ (3$\sigma$) are only allowed if the atmosphere lacks any strong absorbers across the mid-IR wavelength range, a scenario that we deem unlikely. We additionally model the emission spectra for bare-rock planets of various compositions. We find that a variety of silicate surfaces match the measured eclipse depth to within 1$\sigma$, and the best-fit grey albedo is $0.02 \pm 0.11$. We conclude that planned secondary eclipse observations at 12.8 $\mu$m will serve to validate the high observed brightness temperature of TRAPPIST-1 b, but are unlikely to further distinguish among the consistent atmospheric and bare-rock scenarios., Comment: 8 pages, 4 figures, submitted to ApJL