Search

This paper presents a comprehensive study on the scalability challenges and opportunities in quantum communication networks, with the goal of determining parameters that impact networks most as well as the trends that appear when scaling networks. We design simulations of quantum networks comprised of router nodes made up of trapped-ion qubits, separated by quantum repeaters in the form of Bell State Measurement (BSM) nodes. Such networks hold the promise of securely sharing quantum information and enabling high-power distributed quantum computing. Despite the promises, quantum networks encounter scalability issues due to noise and operational errors. Through a modular approach, our research aims to surmount these challenges, focusing on effects from scaling node counts and separation distances while monitoring low-quality communication arising from decoherence effects. We aim to pinpoint the critical features within networks essential for advancing scalable, large-scale quantum computing systems. Our findings underscore the impact of several network parameters on scalability, highlighting a critical insight into the trade-offs between the number of repeaters and the quality of entanglement generated. This paper lays the groundwork for future explorations into optimized quantum network designs and protocols., Comment: 10 pages, 11 figures

We find that the cross-entropy loss curves of neural language models empirically adhere to a scaling law with learning rate (LR) annealing over training steps ($s$): $$L(s) = L_0 + A\cdot S_1^{-\alpha} - C\cdot S_2$$ Where $S_1$ is forward area and $S_2$ is learning rate annealing area. This formulation takes into account two factors: (1) The forward scaling defined as typical scaling law, and (2) the additional loss drop brought by LR annealing. Therefore, this formulation can describe the full loss curve at each step, rather than the single loss point at the end of training. Applying the scaling law with LR annealing and fitting only one or two training curves, we can accurately predict the loss of language model training at any given step and across any learning rate scheduler (LRS). Furthermore, this equation accurately describes the dynamics during training process, and provides a theoretical verification and explanation for numerous experimental findings of previous studies, particularly those focusing on LR schedule and LR annealing. The resulting insights, also serve as a guide for researchers to select critical LRS in advance by prediction using our equation. Most significantly, since all the points in a full training curve follow the equation, we can achieve accurate loss prediction at any given step across any learning rate scheduler, while expending less than 1\% of the computational cost required by the chinchilla scaling law to fit language modeling loss. This approach extremely democratizes scaling law fitting and predicting in developing large language models., Comment: 25 pages, 23 figures

This research introduces the Multilevel Embedding Association Test (ML-EAT), a method designed for interpretable and transparent measurement of intrinsic bias in language technologies. The ML-EAT addresses issues of ambiguity and difficulty in interpreting the traditional EAT measurement by quantifying bias at three levels of increasing granularity: the differential association between two target concepts with two attribute concepts; the individual effect size of each target concept with two attribute concepts; and the association between each individual target concept and each individual attribute concept. Using the ML-EAT, this research defines a taxonomy of EAT patterns describing the nine possible outcomes of an embedding association test, each of which is associated with a unique EAT-Map, a novel four-quadrant visualization for interpreting the ML-EAT. Empirical analysis of static and diachronic word embeddings, GPT-2 language models, and a CLIP language-and-image model shows that EAT patterns add otherwise unobservable information about the component biases that make up an EAT; reveal the effects of prompting in zero-shot models; and can also identify situations when cosine similarity is an ineffective metric, rendering an EAT unreliable. Our work contributes a method for rendering bias more observable and interpretable, improving the transparency of computational investigations into human minds and societies., Comment: Accepted at Artificial Intelligence, Ethics, and Society 2024

Popular and news media often portray teenagers with sensationalism, as both a risk to society and at risk from society. As AI begins to absorb some of the epistemic functions of traditional media, we study how teenagers in two countries speaking two languages: 1) are depicted by AI, and 2) how they would prefer to be depicted. Specifically, we study the biases about teenagers learned by static word embeddings (SWEs) and generative language models (GLMs), comparing these with the perspectives of adolescents living in the U.S. and Nepal. We find English-language SWEs associate teenagers with societal problems, and more than 50% of the 1,000 words most associated with teenagers in the pretrained GloVe SWE reflect such problems. Given prompts about teenagers, 30% of outputs from GPT2-XL and 29% from LLaMA-2-7B GLMs discuss societal problems, most commonly violence, but also drug use, mental illness, and sexual taboo. Nepali models, while not free of such associations, are less dominated by social problems. Data from workshops with N=13 U.S. adolescents and N=18 Nepalese adolescents show that AI presentations are disconnected from teenage life, which revolves around activities like school and friendship. Participant ratings of how well 20 trait words describe teens are decorrelated from SWE associations, with Pearson's r=.02, n.s. in English FastText and r=.06, n.s. in GloVe; and r=.06, n.s. in Nepali FastText and r=-.23, n.s. in GloVe. U.S. participants suggested AI could fairly present teens by highlighting diversity, while Nepalese participants centered positivity. Participants were optimistic that, if it learned from adolescents, rather than media sources, AI could help mitigate stereotypes. Our work offers an understanding of the ways SWEs and GLMs misrepresent a developmentally vulnerable group and provides a template for less sensationalized characterization., Comment: Accepted at Artificial Intelligence, Ethics, and Society 2024

Multimodal AI models capable of associating images and text hold promise for numerous domains, ranging from automated image captioning to accessibility applications for blind and low-vision users. However, uncertainty about bias has in some cases limited their adoption and availability. In the present work, we study 43 CLIP vision-language models to determine whether they learn human-like facial impression biases, and we find evidence that such biases are reflected across three distinct CLIP model families. We show for the first time that the the degree to which a bias is shared across a society predicts the degree to which it is reflected in a CLIP model. Human-like impressions of visually unobservable attributes, like trustworthiness and sexuality, emerge only in models trained on the largest dataset, indicating that a better fit to uncurated cultural data results in the reproduction of increasingly subtle social biases. Moreover, we use a hierarchical clustering approach to show that dataset size predicts the extent to which the underlying structure of facial impression bias resembles that of facial impression bias in humans. Finally, we show that Stable Diffusion models employing CLIP as a text encoder learn facial impression biases, and that these biases intersect with racial biases in Stable Diffusion XL-Turbo. While pretrained CLIP models may prove useful for scientific studies of bias, they will also require significant dataset curation when intended for use as general-purpose models in a zero-shot setting., Comment: Accepted at Artificial Intelligence, Ethics, and Society 2024

Equitable urban transportation applications require high-fidelity digital representations of the built environment: not just streets and sidewalks, but bike lanes, marked and unmarked crossings, curb ramps and cuts, obstructions, traffic signals, signage, street markings, potholes, and more. Direct inspections and manual annotations are prohibitively expensive at scale. Conventional machine learning methods require substantial annotated training data for adequate performance. In this paper, we consider vision language models as a mechanism for annotating diverse urban features from satellite images, reducing the dependence on human annotation to produce large training sets. While these models have achieved impressive results in describing common objects in images captured from a human perspective, their training sets are less likely to include strong signals for esoteric features in the built environment, and their performance in these settings is therefore unclear. We demonstrate proof-of-concept combining a state-of-the-art vision language model and variants of a prompting strategy that asks the model to consider segmented elements independently of the original image. Experiments on two urban features -- stop lines and raised tables -- show that while direct zero-shot prompting correctly annotates nearly zero images, the pre-segmentation strategies can annotate images with near 40% intersection-over-union accuracy. We describe how these results inform a new research agenda in automatic annotation of the built environment to improve equity, accessibility, and safety at broad scale and in diverse environments.

Abstention, the refusal of large language models (LLMs) to provide an answer, is increasingly recognized for its potential to mitigate hallucinations and enhance safety in LLM systems. In this survey, we introduce a framework to examine abstention from three perspectives: the query, the model, and human values. We organize the literature on abstention methods, benchmarks, and evaluation metrics using this framework, and discuss merits and limitations of prior work. We further identify and motivate areas for future work, centered around whether abstention can be achieved as a meta-capability that transcends specific tasks or domains, while still providing opportunities to optimize abstention abilities based on context., Comment: preprint

Language model capabilities predictably improve from scaling a model's size and training data. Motivated by this, increasingly large language models have been trained, yielding an array of impressive capabilities. Yet these models are vulnerable to adversarial prompts, such as "jailbreaks" that hijack models to perform undesired behaviors, posing a significant risk of misuse. Prior work indicates that computer vision models become more robust with model and data scaling, raising the question: does language model robustness also improve with scale? We study this question empirically, finding that larger models respond substantially better to adversarial training, but there is little to no benefit from model scale in the absence of explicit defenses., Comment: 31 pages; edit fixed metadata typo (author name)

Applications to support pedestrian mobility in urban areas require a complete, and routable graph representation of the built environment. Globally available information, including aerial imagery provides a scalable source for constructing these path networks, but the associated learning problem is challenging: Relative to road network pathways, pedestrian network pathways are narrower, more frequently disconnected, often visually and materially variable in smaller areas, and their boundaries are broken up by driveway incursions, alleyways, marked or unmarked crossings through roadways. Existing algorithms to extract pedestrian pathway network graphs are inconsistently evaluated and tend to ignore routability, making it difficult to assess utility for mobility applications: Even if all path segments are available, discontinuities could dramatically and arbitrarily shift the overall path taken by a pedestrian. In this paper, we describe a first standard benchmark for the pedestrian pathway graph extraction problem, comprising the largest available dataset equipped with manually vetted ground truth annotations (covering $3,000 km^2$ land area in regions from 8 cities), and a family of evaluation metrics centering routability and downstream utility. By partitioning the data into polygons at the scale of individual intersections, we compute local routability as an efficient proxy for global routability. We consider multiple measures of polygon-level routability and compare predicted measures with ground truth to construct evaluation metrics. Using these metrics, we show that this benchmark can surface strengths and weaknesses of existing methods that are hidden by simple edge-counting metrics over single-region datasets used in prior work, representing a challenging, high-impact problem in computer vision and machine learning., Comment: arXiv admin note: text overlap with arXiv:2303.02323

The ongoing efforts to optimize Li-ion batteries led to the interest in intercalation of nanoscale layered compounds, including bilayer graphene. Its lithium intercalation has been demonstrated recently but the mechanisms underpinning the storage capacity remain poorly understood. Here, using magnetotransport measurements, we report in-operando intercalation dynamics of bilayer graphene. Unexpectedly, we find four distinct intercalation stages that correspond to well-defined Li-ion densities. We refer to these stages as 'in-plane', with no in-plane analogues in bulk graphite. The fully intercalated bilayers represent a stoichiometric compound C14LiC14 with a Li density of 2.7x10^{14} cm^{-2}, notably lower than fully intercalated graphite. Combining the experimental findings and DFT calculations, we show that the critical step in bilayer intercalation is a transition from AB to AA stacking which occurs at a density of 0.9x10^{14} cm^{-2}. Our findings reveal the mechanism and limits for electrochemical intercalation of bilayer graphene and suggest possible avenues for increasing the Li storage capacity., Comment: 30 pages, 17 figures

Tendon-driven robotic catheters are capable of precise execution of minimally invasive cardiac procedures including ablations and imaging. These procedures require accurate mathematical models of not only the catheter and tendons but also their interactions with surrounding tissue and vasculature in order to control the robot path and interaction. This paper presents a mechanical model of a tendon-driven robotic catheter system based on Cosserat rods and integrated with a stable, implicit Euler scheme. We implement the Cosserat rod as a model for a simple catheter centerline and validate its physical accuracy against a large deformation analytical model and experimental data. The catheter model is then supplemented by adding a second Cosserat rod to model a single tendon, using penalty forces to define the constraints of the tendon-catheter system. All the model parameters are defined by the catheter properties established by the design. The combined model is validated against experimental data to confirm its physical accuracy. This model represents a new contribution to the field of robotic catheter modeling in which both the tendons and catheter are modeled by mechanical Cosserat rods and fully-validated against experimental data in the case of the single rod system., Comment: 24 pages, 23 figures

We present an algorithm that, for every fixed degree $n\ge 3$, will enumerate all degree-$n$ places of the projective line over a finite field $k$ up to the natural action of $\operatorname{PGL}_2(k)$ using $O(\log q)$ space and $\widetilde{O}(q^{n-3})$ time, where $q=\#k$. Since there are $\Theta(q^{n-3})$orbits of $\operatorname{PGL}_2(k)$ acting on the set of degree-$n$ places, the algorithm is quasilinear in the size of its output. The algorithm is probabilistic unless we assume the extended Riemann hypothesis, because its complexity depends on that of polynomial factorization: for odd $n$, it involves factoring $\Theta(q^{n-3})$ polynomials over $k$ of degree up to ${1 + ((n-1)/2)^n}$. For composite degrees $n$, earlier work of the author gives an algorithm for enumerating $\operatorname{PGL}_2(k)$ orbit representatives for degree-$n$ places of $\mathbf{P}^1$ over $k$ that runs in time $\widetilde{O}(q^{n-3})$ independent of the extended Riemann hypothesis, but that uses $O(q^{n-3})$ space. We also present an algorithm for enumerating orbit representatives for the action of $\operatorname{PGL}_2(k)$ on the degree-$n$ effective divisors of $\mathbf{P}^1$ over finite fields $k$. The two algorithms depend on one another; our method of enumerating orbits of places of odd degree $n$ depends on enumerating orbits of effective divisors of degree $(n+1)/2$., Comment: 23 pages

We present the first detailed study of the effect of a strong magnetic field on single-electron pumping in a device utilising a finger-gate split-gate configuration. In the quantum Hall regime, we demonstrate electron pumping from Landau levels in the leads, where the measurements exhibit pronounced oscillations in the lengths of the pumping plateaus with the magnetic field, reminiscent of Shubnikov-de Haas oscillations. This similarity indicates that the pumping process is dependent on the density of states of the 2D electron gas over a narrow energy window. Based on these observations, we develop a new theoretical description of the operation of single-electron pumps which for the first time allows for the determination of the physical parameters of the experiment; such as the capture energy of the electrons, the broadening of the quantised Landau levels in the leads, and the quantum lifetime of the electrons., Comment: 8 pages, 4 figures

In this paper, we introduce and evaluate a tool for researchers and practitioners to assess the actionability of information provided to users to support algorithmic recourse. While there are clear benefits of recourse from the user's perspective, the notion of actionability in explainable AI research remains vague, and claims of `actionable' explainability techniques are based on the researchers' intuition. Inspired by definitions and instruments for assessing actionability in other domains, we construct a seven-question tool and evaluate its effectiveness through two user studies. We show that the tool discriminates actionability across explanation types and that the distinctions align with human judgements. We show the impact of context on actionability assessments, suggesting that domain-specific tool adaptations may foster more human-centred algorithmic systems. This is a significant step forward for research and practices into actionable explainability and algorithmic recourse, providing the first clear human-centred definition and tool for assessing actionability in explainable AI., Comment: 10 pages, 4 figures

A novel method to reconstruct the energy of hadronic showers in the CMS High Granularity Calorimeter (HGCAL) is presented. The HGCAL is a sampling calorimeter with very fine transverse and longitudinal granularity. The active media are silicon sensors and scintillator tiles readout by SiPMs and the absorbers are a combination of lead and Cu/CuW in the electromagnetic section, and steel in the hadronic section. The shower reconstruction method is based on graph neural networks and it makes use of a dynamic reduction network architecture. It is shown that the algorithm is able to capture and mitigate the main effects that normally hinder the reconstruction of hadronic showers using classical reconstruction methods, by compensating for fluctuations in the multiplicity, energy, and spatial distributions of the shower's constituents. The performance of the algorithm is evaluated using test beam data collected in 2018 prototype of the CMS HGCAL accompanied by a section of the CALICE AHCAL prototype. The capability of the method to mitigate the impact of energy leakage from the calorimeter is also demonstrated., Comment: Prepared for submission to JINST

The rapid proliferation of generative AI has raised questions about the competitiveness of lower-parameter, locally tunable, open-weight models relative to high-parameter, API-guarded, closed-weight models in terms of performance, domain adaptation, cost, and generalization. Centering under-resourced yet risk-intolerant settings in government, research, and healthcare, we see for-profit closed-weight models as incompatible with requirements for transparency, privacy, adaptability, and standards of evidence. Yet the performance penalty in using open-weight models, especially in low-data and low-resource settings, is unclear. We assess the feasibility of using smaller, open-weight models to replace GPT-4-Turbo in zero-shot, few-shot, and fine-tuned regimes, assuming access to only a single, low-cost GPU. We assess value-sensitive issues around bias, privacy, and abstention on three additional tasks relevant to those topics. We find that with relatively low effort, very low absolute monetary cost, and relatively little data for fine-tuning, small open-weight models can achieve competitive performance in domain-adapted tasks without sacrificing generality. We then run experiments considering practical issues in bias, privacy, and hallucination risk, finding that open models offer several benefits over closed models. We intend this work as a case study in understanding the opportunity cost of reproducibility and transparency over for-profit state-of-the-art zero shot performance, finding this cost to be marginal under realistic settings., Comment: Accepted at the ACM Conference on Fairness, Accountability, and Transparency (FAccT) 2024

Future space missions that aim to detect and characterize Earth-like exoplanets will require an instrument that efficiently measures spectra of these planets, placing strict requirements on detector performance. The upcoming Roman Space Telescope will demonstrate the performance of an electron-multiplying charge-coupled device (EMCCD) as part of the coronagraphic instrument (CGI). The recent LUVOIR and HabEx studies baselined pairing such a detector with an integral field spectrograph (IFS) to take spectra of multiple exoplanets and debris disks simultaneously. We investigate the scientific impact of a noiseless energy-resolving detector for the planned Habitable Worlds Observatory's (HWO) coronagraphic instrument. By assuming higher quantum efficiency, higher optical throughput, and zero noise, we effectively place upper limits on the impact of advancing detector technologies. We find that energy-resolving detectors would potentially take spectra of hundreds of additional exoplanets "for free" over the course of an HWO survey, greatly increasing its scientific yield., Comment: 29 pages, 2 figures, 2 tables, Accepted to JATIS

Hexagonal boron nitride (h-BN) is brittle, however, its atomic-scale structural engineering can lead to unprecedented physical properties. Here we report the bulk synthesis of high-density crystalline h-BN solids by using high-temperature spark plasma sintering (SPS) of micron size h-BN powders. In addition to the high mechanical strength and ductile response of such materials, we have obtained anomalous values of dielectric constant beyond theoretical limits, high thermal conductivity, and exceptional neutron radiation shielding capability. Through exhaustive characterizations we reveal that SPS induces non-basal plane crystallinity, twisting of layers, and facilitates inter-grain fusion with a high degree of in-plane alignment across macroscale dimensions, resulting in near-theoretical density and anomalous properties. Our findings highlight the importance of material design, via new approaches such as twisting and interconnections between atomically thin layers, to create novel ceramics with properties that could go beyond their intrinsic theoretical predictions., Comment: Authors revised version, 46 pages, 4 figures

The correct model response in the face of uncertainty is to abstain from answering a question so as not to mislead the user. In this work, we study the ability of LLMs to abstain from answering context-dependent science questions when provided insufficient or incorrect context. We probe model sensitivity in several settings: removing gold context, replacing gold context with irrelevant context, and providing additional context beyond what is given. In experiments on four QA datasets with four LLMs, we show that performance varies greatly across models, across the type of context provided, and also by question type; in particular, many LLMs seem unable to abstain from answering boolean questions using standard QA prompts. Our analysis also highlights the unexpected impact of abstention performance on QA task accuracy. Counter-intuitively, in some settings, replacing gold context with irrelevant context or adding irrelevant context to gold context can improve abstention performance in a way that results in improvements in task performance. Our results imply that changes are needed in QA dataset design and evaluation to more effectively assess the correctness and downstream impacts of model abstention.

A Kinetic Inductance Traveling Wave amplifier (KIT) utilizes the nonlinear kinetic inductance of superconducting films, particularly Niobium Titanium Nitride (NbTiN), for parametric amplification. These amplifiers achieve remarkable performance in terms of gain, bandwidth, compression power, and frequently approach the quantum limit for noise. However, most KIT demonstrations have been isolated from practical device readout systems. Using a KIT as the first amplifier in the readout chain of an unoptimized microwave SQUID multiplexer coupled to a transition-edge sensor microcalorimeter we see an initial improvement in the flux noise. One challenge in KIT integration is the considerable microwave pump power required to drive the non-linearity. To address this, we have initiated efforts to reduce the pump power by using thinner NbTiN films and an inverted microstrip transmission line design. In this article, we present the new transmission line design, fabrication procedure, and initial device characterization -- including gain and added noise. These devices exhibit over 10 dB of gain with a 3 dB bandwidth of approximately 5.5-7.25 GHz, a maximum practical gain of 12 dB and typical gain ripple under 4 dB peak-to-peak. We observe an appreciable impedance mismatch in the NbTiN transmission line, which is likely the source of the majority of the gain ripple. Finally we perform an initial noise characterization and demonstrate system-added noise of three quanta or less over nearly the entire 3 dB bandwidth.

We introduce the Visual Experience Dataset (VEDB), a compilation of over 240 hours of egocentric video combined with gaze- and head-tracking data that offers an unprecedented view of the visual world as experienced by human observers. The dataset consists of 717 sessions, recorded by 58 observers ranging from 6-49 years old. This paper outlines the data collection, processing, and labeling protocols undertaken to ensure a representative sample and discusses the potential sources of error or bias within the dataset. The VEDB's potential applications are vast, including improving gaze tracking methodologies, assessing spatiotemporal image statistics, and refining deep neural networks for scene and activity recognition. The VEDB is accessible through established open science platforms and is intended to be a living dataset with plans for expansion and community contributions. It is released with an emphasis on ethical considerations, such as participant privacy and the mitigation of potential biases. By providing a dataset grounded in real-world experiences and accompanied by extensive metadata and supporting code, the authors invite the research community to utilize and contribute to the VEDB, facilitating a richer understanding of visual perception and behavior in naturalistic settings., Comment: 40 pages, 1 table, 9 figures

Let $C$ and $C'$ be curves over a finite field $K$, provided with embeddings $\iota$ and $\iota'$ into their Jacobian varieties. Let $D\to C$ and $D'\to C'$ be the pullbacks (via these embeddings) of the multiplication-by-$2$ maps on the Jacobians. We say that $(C,\iota)$ and $(C',\iota')$ are \emph{doubly isogenous} if $\mathrm{Jac}(C)$ and $\mathrm{Jac}(C')$ are isogenous over $K$ and $\mathrm{Jac}(D)$ and $\mathrm{Jac}(D')$ are isogenous over~$K$. When we restrict attention to the case where $C$ and $C'$ are curves of genus $2$ whose groups of $K$-rational automorphisms are isomorphic to the dihedral group $D_6$ of order $12$, we find many more doubly isogenous pairs than one would expect from reasonable heuristics. Our analysis of this overabundance of doubly isogenous curves over finite fields leads to the construction of a pair of doubly isogenous curves over a number field. That such a global example exists seems extremely surprising. We show that the Zilber--Pink conjecture implies that there can only be finitely many such examples. When we exclude reductions of this pair of global curves in our counts, we find that the data for the remaining curves is consistent with our original heuristic. Computationally, we find that doubly isogenous curves in our family of $D_6$ curves can be distinguished from one another by considering the isogeny classes of the Prym varieties of certain unramified covers of exponent $3$ and $4$. We discuss how our family of curves can be potentially be used to obtain a deterministic polynomial-time algorithm to factor univariate polynomials over finite fields via an argument of Kayal and Poonen., Comment: 40 pages

Radiative transfer (RT) models are critical in the interpretation of exoplanetary spectra, in simulating exoplanet climates and when designing the specifications of future flagship observatories. However, most models differ in methodologies and input data, which can lead to significantly different spectra. In this paper, we present the experimental protocol of the MALBEC (Modeling Atmospheric Lines By the Exoplanet Community) project. MALBEC is an exoplanet model intercomparison project (exoMIP) that belongs to the CUISINES (Climates Using Interactive Suites of Intercomparisons Nested for Exoplanet Studies) framework which aims to provide the exoplanet community with a large and diverse set of comparison and validation of models. The proposed protocol tests include a large set of initial participating RT models, a broad range of atmospheres (from Hot Jupiters to temperate terrestrials) and several observation geometries, which would allow us to quantify and compare the differences between different RT models used by the exoplanetary community. Two types of tests are proposed: transit spectroscopy and direct imaging modeling, with results from the proposed tests to be published in dedicated follow-up papers. To encourage the community to join this comparison effort and as an example, we present simulation results for one specific transit case (GJ-1214 b), in which we find notable differences in how the various codes handle the discretization of the atmospheres (e.g., sub-layering), the treatment of molecular opacities (e.g., correlated-k, line-by-line) and the default spectroscopic repositories generally used by each model (e.g., HITRAN, HITEMP, ExoMol).

We present an algorithm that, for every fixed genus $g$, will enumerate all hyperelliptic curves of genus $g$ over a finite field $k$ of odd characteristic in quasilinear time; that is, the time required for the algorithm is $\widetilde{O}(q^{2g-1})$, where $q=\#k$. Such an algorithm already exists in the case $g=2$, thanks to work of Mestre and Cardona and Quer, and in the case $g=3$, thanks to work of Lercier and Ritzenthaler. Experimentally, it appears that our new algorithm is about two orders of magnitude faster in practice than ones based on their work., Comment: 23 pages. Corrected small typos, added a reference to a paper of Nart, and added links to Internet Archive library copies of two books