Showing 1–50 of 69 results for author: Hudson, T

The Power of Next-Frame Prediction for Learning Physical Laws

Authors: Thomas Winterbottom, G. Thomas Hudson, Daniel Kluvanec, Dean Slack, Jamie Sterling, Junjie Shentu, Chenghao Xiao, Zheming Zhou, Noura Al Moubayed

Abstract: Next-frame prediction is a useful and powerful method for modelling and understanding the dynamics of video data. Inspired by the empirical success of causal language modelling and next-token prediction in language modelling, we explore the extent to which next-frame prediction serves as a strong foundational learning strategy (analogous to language modelling) for inducing an understanding of the… ▽ More Next-frame prediction is a useful and powerful method for modelling and understanding the dynamics of video data. Inspired by the empirical success of causal language modelling and next-token prediction in language modelling, we explore the extent to which next-frame prediction serves as a strong foundational learning strategy (analogous to language modelling) for inducing an understanding of the visual world. In order to quantify the specific visual understanding induced by next-frame prediction, we introduce six diagnostic simulation video datasets derived from fundamental physical laws created by varying physical constants such as gravity and mass. We demonstrate that our models trained only on next-frame prediction are capable of predicting the value of these physical constants (e.g. gravity) without having been trained directly to learn these constants via a regression task. We find that the generative training phase alone induces a model state that can predict physical constants significantly better than that of a random model, improving the loss by a factor of between 1.28 to 6.24. We conclude that next-frame prediction shows great promise as a general learning strategy to induce understanding of the many `laws' that govern the visual domain without the need for explicit labelling. △ Less

Submitted 21 May, 2024; originally announced May 2024.

Comments: 7 Figures, 12 Pages, 1 Table

MSC Class: 68T45 ACM Class: I.2.6; I.2.10

RAR-b: Reasoning as Retrieval Benchmark

Authors: Chenghao Xiao, G Thomas Hudson, Noura Al Moubayed

Abstract: Semantic textual similartiy (STS) and information retrieval tasks (IR) tasks have been the two major avenues to record the progress of embedding models in the past few years. Under the emerging Retrieval-augmented Generation (RAG) paradigm, we envision the need to evaluate next-level language understanding abilities of embedding models, and take a conscious look at the reasoning abilities stored i… ▽ More Semantic textual similartiy (STS) and information retrieval tasks (IR) tasks have been the two major avenues to record the progress of embedding models in the past few years. Under the emerging Retrieval-augmented Generation (RAG) paradigm, we envision the need to evaluate next-level language understanding abilities of embedding models, and take a conscious look at the reasoning abilities stored in them. Addressing this, we pose the question: Can retrievers solve reasoning problems? By transforming reasoning tasks into retrieval tasks, we find that without specifically trained for reasoning-level language understanding, current state-of-the-art retriever models may still be far from being competent for playing the role of assisting LLMs, especially in reasoning-intensive tasks. Moreover, albeit trained to be aware of instructions, instruction-aware IR models are often better off without instructions in inference time for reasoning tasks, posing an overlooked retriever-LLM behavioral gap for the research community to align. However, recent decoder-based embedding models show great promise in narrowing the gap, highlighting the pathway for embedding models to achieve reasoning-level language understanding. We also show that, although current off-the-shelf re-ranker models fail on these tasks, injecting reasoning abilities into them through fine-tuning still appears easier than doing so to bi-encoders, and we are able to achieve state-of-the-art performance across all tasks by fine-tuning a reranking model. We release Reasoning as Retrieval Benchmark (RAR-b), a holistic suite of tasks and settings to evaluate the reasoning abilities stored in retriever models. RAR-b is available at https://github.com/gowitheflow-1998/RAR-b. △ Less

Submitted 12 May, 2024; v1 submitted 9 April, 2024; originally announced April 2024.

Comments: v2, small typo fixes

Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context

Authors: Gemini Team, Petko Georgiev, Ving Ian Lei, Ryan Burnell, Libin Bai, Anmol Gulati, Garrett Tanzer, Damien Vincent, Zhufeng Pan, Shibo Wang, Soroosh Mariooryad, Yifan Ding, Xinyang Geng, Fred Alcober, Roy Frostig, Mark Omernick, Lexi Walker, Cosmin Paduraru, Christina Sorokin, Andrea Tacchetti, Colin Gaffney, Samira Daruki, Olcan Sercinoglu, Zach Gleicher, Juliette Love , et al. (1092 additional authors not shown)

Abstract: 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… ▽ More 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. △ Less

Submitted 14 June, 2024; v1 submitted 8 March, 2024; originally announced March 2024.

Pixel Sentence Representation Learning

Authors: Chenghao Xiao, Zhuoxu Huang, Danlu Chen, G Thomas Hudson, Yizhi Li, Haoran Duan, Chenghua Lin, Jie Fu, Jungong Han, Noura Al Moubayed

Abstract: Pretrained language models are long known to be subpar in capturing sentence and document-level semantics. Though heavily investigated, transferring perturbation-based methods from unsupervised visual representation learning to NLP remains an unsolved problem. This is largely due to the discreteness of subword units brought by tokenization of language models, limiting small perturbations of inputs… ▽ More Pretrained language models are long known to be subpar in capturing sentence and document-level semantics. Though heavily investigated, transferring perturbation-based methods from unsupervised visual representation learning to NLP remains an unsolved problem. This is largely due to the discreteness of subword units brought by tokenization of language models, limiting small perturbations of inputs to form semantics-preserved positive pairs. In this work, we conceptualize the learning of sentence-level textual semantics as a visual representation learning process. Drawing from cognitive and linguistic sciences, we introduce an unsupervised visual sentence representation learning framework, employing visually-grounded text perturbation methods like typos and word order shuffling, resonating with human cognitive patterns, and enabling perturbation to texts to be perceived as continuous. Our approach is further bolstered by large-scale unsupervised topical alignment training and natural language inference supervision, achieving comparable performance in semantic textual similarity (STS) to existing state-of-the-art NLP methods. Additionally, we unveil our method's inherent zero-shot cross-lingual transferability and a unique leapfrogging pattern across languages during iterative training. To our knowledge, this is the first representation learning method devoid of traditional language models for understanding sentence and document semantics, marking a stride closer to human-like textual comprehension. Our code is available at https://github.com/gowitheflow-1998/Pixel-Linguist △ Less

Submitted 12 February, 2024; originally announced February 2024.

Gemini: A Family of Highly Capable Multimodal Models

Authors: Gemini Team, Rohan Anil, Sebastian Borgeaud, Jean-Baptiste Alayrac, Jiahui Yu, Radu Soricut, Johan Schalkwyk, Andrew M. Dai, Anja Hauth, Katie Millican, David Silver, Melvin Johnson, Ioannis Antonoglou, Julian Schrittwieser, Amelia Glaese, Jilin Chen, Emily Pitler, Timothy Lillicrap, Angeliki Lazaridou, Orhan Firat, James Molloy, Michael Isard, Paul R. Barham, Tom Hennigan, Benjamin Lee , et al. (1325 additional authors not shown)

Abstract: This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultr… ▽ More This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. We believe that the new capabilities of the Gemini family in cross-modal reasoning and language understanding will enable a wide variety of use cases. We discuss our approach toward post-training and deploying Gemini models responsibly to users through services including Gemini, Gemini Advanced, Google AI Studio, and Cloud Vertex AI. △ Less

Submitted 17 June, 2024; v1 submitted 18 December, 2023; originally announced December 2023.

A geometric $C_2$-equivariant Bézout Theorem

Authors: Steven R. Costenoble, Thomas Hudson

Abstract: Classically, Bézout's theorem says that an intersection of hypersurfaces in a projective space is rationally equivalent to a number of copies of a smaller projective space, the number depending on the degrees of the hypersurfaces. We give a generalization of that result to the context of $C_2$-equivariant hypersurfaces in $C_2$-equivariant linear projective space, expressing the intersection as a… ▽ More Classically, Bézout's theorem says that an intersection of hypersurfaces in a projective space is rationally equivalent to a number of copies of a smaller projective space, the number depending on the degrees of the hypersurfaces. We give a generalization of that result to the context of $C_2$-equivariant hypersurfaces in $C_2$-equivariant linear projective space, expressing the intersection as a linear combination of equivariant Schubert varieties. △ Less

Submitted 1 December, 2023; originally announced December 2023.

Comments: 45 pages

MSC Class: Primary: 55N91; 14N10; Secondary: 14M15; 14N15; 55R40; 55R91

Navigating to Success in Multi-Modal Human-Robot Collaboration: Analysis and Corpus Release

Authors: Stephanie M. Lukin, Kimberly A. Pollard, Claire Bonial, Taylor Hudson, Ron Arstein, Clare Voss, David Traum

Abstract: Human-guided robotic exploration is a useful approach to gathering information at remote locations, especially those that might be too risky, inhospitable, or inaccessible for humans. Maintaining common ground between the remotely-located partners is a challenge, one that can be facilitated by multi-modal communication. In this paper, we explore how participants utilized multiple modalities to inv… ▽ More Human-guided robotic exploration is a useful approach to gathering information at remote locations, especially those that might be too risky, inhospitable, or inaccessible for humans. Maintaining common ground between the remotely-located partners is a challenge, one that can be facilitated by multi-modal communication. In this paper, we explore how participants utilized multiple modalities to investigate a remote location with the help of a robotic partner. Participants issued spoken natural language instructions and received from the robot: text-based feedback, continuous 2D LIDAR map**, and upon-request static photographs. We noticed that different strategies were adopted in terms of use of the modalities, and hypothesize that these differences may be correlated with success at several exploration sub-tasks. We found that requesting photos may have improved the identification and counting of some key entities (doorways in particular) and that this strategy did not hinder the amount of overall area exploration. Future work with larger samples may reveal the effects of more nuanced photo and dialogue strategies, which can inform the training of robotic agents. Additionally, we announce the release of our unique multi-modal corpus of human-robot communication in an exploration context: SCOUT, the Situated Corpus on Understanding Transactions. △ Less

Submitted 26 October, 2023; originally announced October 2023.

Comments: 7 pages, 3 figures

Journal ref: Proceedings of the 2023 IEEE Robot and Human Interactive Communication Conference

Length is a Curse and a Blessing for Document-level Semantics

Authors: Chenghao Xiao, Yizhi Li, G Thomas Hudson, Chenghua Lin, Noura Al Moubayed

Abstract: In recent years, contrastive learning (CL) has been extensively utilized to recover sentence and document-level encoding capability from pre-trained language models. In this work, we question the length generalizability of CL-based models, i.e., their vulnerability towards length-induced semantic shift. We verify not only that length vulnerability is a significant yet overlooked research gap, but… ▽ More In recent years, contrastive learning (CL) has been extensively utilized to recover sentence and document-level encoding capability from pre-trained language models. In this work, we question the length generalizability of CL-based models, i.e., their vulnerability towards length-induced semantic shift. We verify not only that length vulnerability is a significant yet overlooked research gap, but we can devise unsupervised CL methods solely depending on the semantic signal provided by document length. We first derive the theoretical foundations underlying length attacks, showing that elongating a document would intensify the high intra-document similarity that is already brought by CL. Moreover, we found that isotropy promised by CL is highly dependent on the length range of text exposed in training. Inspired by these findings, we introduce a simple yet universal document representation learning framework, LA(SER)$^{3}$: length-agnostic self-reference for semantically robust sentence representation learning, achieving state-of-the-art unsupervised performance on the standard information retrieval benchmark. △ Less

Submitted 24 October, 2023; originally announced October 2023.

Comments: Accepted at EMNLP 2023. Our code is publicly available at https://github.com/gowitheflow-1998/LA-SER-cubed

Dislocation dynamics in Ni-based superalloys: Parameterising dislocation trajectories from atomistic simulations

Authors: Geraldine Anis, Thomas Hudson, Peter Brommer

Abstract: Nanoscale precipitates in the microstructure of nickel-based superalloys hinder dislocation motion, which results in an extraordinary strengthening effect at elevated temperatures. We used molecular dynamics (MD) with classical effective potential to observe the movement of an $\frac{a}{2}\langle110\rangle\{111\}$ edge dislocation under shear in pure Ni, which represents the Ni solid solution matr… ▽ More Nanoscale precipitates in the microstructure of nickel-based superalloys hinder dislocation motion, which results in an extraordinary strengthening effect at elevated temperatures. We used molecular dynamics (MD) with classical effective potential to observe the movement of an $\frac{a}{2}\langle110\rangle\{111\}$ edge dislocation under shear in pure Ni, which represents the Ni solid solution matrix, and extracted the locations of the dislocations. We show how a Differential Evolution Monte Carlo (DE-MC) analysis is an effective way to find the parameters of an equation of motion for the dislocation lines with quantified uncertainties. The parameters of interest were the effective mass, drag coefficient, and force experienced by the dislocation. The marginal parameter and joint posterior distributions were estimated from the accepted samples produced by the DE-MC algorithm. The equation of motion and parameter distributions were used to predict the dislocation positions and velocities at the simulation timesteps, and the mean fit was found to match the MD trajectories with a root mean square error (RMSE) of \SI{0.2}{\nano\metre}. We also discuss how the selected model can be extended to account for the presence of multiple dislocations as well as dislocation-precipitate interactions. This work serves as the first step towards building a predictive surrogate model that describes the deformation behaviour of Ni-based superalloys. △ Less

Submitted 1 March, 2024; v1 submitted 2 October, 2023; originally announced October 2023.

Comments: 19 pages, 7 figures

DAS-N2N: Machine learning Distributed Acoustic Sensing (DAS) signal denoising without clean data

Authors: Sacha Lapins, Antony Butcher, J. -Michael Kendall, Thomas S. Hudson, Anna L. Stork, Maximilian J. Werner, Jemma Gunning, Alex M. Brisbourne

Abstract: This article presents a weakly supervised machine learning method, which we call DAS-N2N, for suppressing strong random noise in distributed acoustic sensing (DAS) recordings. DAS-N2N requires no manually produced labels (i.e., pre-determined examples of clean event signals or sections of noise) for training and aims to map random noise processes to a chosen summary statistic, such as the distribu… ▽ More This article presents a weakly supervised machine learning method, which we call DAS-N2N, for suppressing strong random noise in distributed acoustic sensing (DAS) recordings. DAS-N2N requires no manually produced labels (i.e., pre-determined examples of clean event signals or sections of noise) for training and aims to map random noise processes to a chosen summary statistic, such as the distribution mean, median or mode, whilst retaining the true underlying signal. This is achieved by splicing (joining together) two fibres hosted within a single optical cable, recording two noisy copies of the same underlying signal corrupted by different independent realizations of random observational noise. A deep learning model can then be trained using only these two noisy copies of the data to produce a near fully-denoised copy. Once the model is trained, only noisy data from a single fibre is required. Using a dataset from a DAS array deployed on the surface of the Rutford Ice Stream in Antarctica, we demonstrate that DAS-N2N greatly suppresses incoherent noise and enhances the signal-to-noise ratios (SNR) of natural microseismic icequake events. We further show that this approach is inherently more efficient and effective than standard stop/pass band and white noise (e.g., Wiener) filtering routines, as well as a comparable self-supervised learning method based on masking individual DAS channels. Our preferred model for this task is lightweight, processing 30 seconds of data recorded at a sampling frequency of 1000 Hz over 985 channels (approx. 1 km of fiber) in $<$1 s. Due to the high noise levels in DAS recordings, efficient data-driven denoising methods, such as DAS-N2N, will prove essential to time-critical DAS earthquake detection, particularly in the case of microseismic monitoring. △ Less

Submitted 24 November, 2023; v1 submitted 17 April, 2023; originally announced April 2023.

Comments: Submitted for publication to Geophysical Journal International. For the purpose of open access, the author(s) has applied a Creative Commons Attribution (CC BY) licence to the Author Accepted Manuscript version arising from this submission

Chow-Witt rings and topology of flag varieties

Authors: Thomas Hudson, Ákos K. Matszangosz, Matthias Wendt

Abstract: The paper computes the Witt-sheaf cohomology rings of partial flag varieties in type A in terms of the Pontryagin classes of the subquotient bundles. The proof is based on a Leray-Hirsch-type theorem for Witt-sheaf cohomology for the maximal rank cases, and a detailed study of cohomology ring presentations and annihilators of characteristic classes for the general case. The computations have conse… ▽ More The paper computes the Witt-sheaf cohomology rings of partial flag varieties in type A in terms of the Pontryagin classes of the subquotient bundles. The proof is based on a Leray-Hirsch-type theorem for Witt-sheaf cohomology for the maximal rank cases, and a detailed study of cohomology ring presentations and annihilators of characteristic classes for the general case. The computations have consequences for the topology of real flag manifolds: we show that all torsion in the integral cohomology is 2-torsion, which was not known in full generality previously. This allows for example to compute the Poincaré polynomials of complete flag varieties for cohomology with twisted integer coefficients. The computations also allow to describe the Chow-Witt rings of flag varieties, and we sketch an enumerative application to counting flags satisfying multiple incidence conditions to given hypersurfaces. △ Less

Submitted 21 February, 2023; originally announced February 2023.

Comments: 58 pages

MSC Class: 14M15; 14F25; 19G12; 57T15

Dynamical properties of coarse-grained linear SDEs

Authors: Thomas Hudson, Xingjie Helen Li

Abstract: Coarse-graining or model reduction is a term describing a range of approaches used to extend the time-scale of molecular simulations by reducing the number of degrees of freedom. In the context of molecular simulation, standard coarse-graining approaches approximate the potential of mean force and use this to drive an effective Markovian model. To gain insight into this process, the simple case of… ▽ More Coarse-graining or model reduction is a term describing a range of approaches used to extend the time-scale of molecular simulations by reducing the number of degrees of freedom. In the context of molecular simulation, standard coarse-graining approaches approximate the potential of mean force and use this to drive an effective Markovian model. To gain insight into this process, the simple case of a quadratic energy is studied in an overdamped setting. A hierarchy of reduced models is derived and analysed, and the merits of these different coarse-graining approaches are discussed. In particular, while standard recipes for model reduction accurately capture static equilibrium statistics, it is shown that dynamical statistics such as the mean-squared displacement display systematic error, even when a system exhibits large time-scale separation. In the linear setting studied, it is demonstrated both analytically and numerically that such models can be augmented in a simple way to better capture dynamical statistics. △ Less

Submitted 13 November, 2023; v1 submitted 13 February, 2023; originally announced February 2023.

Comments: 33 pages, 8 figures. Version accepted for publication in SIAM Multiscale Modeling and Simulation

MSC Class: 60H10; 34F05

An algebraic $C_2$-equivariant Bézout's theorem

Authors: Steven R. Costenoble, Thomas Hudson, Sean Tilson

Abstract: Bézout's theorem, nonequivariantly, can be interpreted as a calculation of the Euler class of a sum of line bundles over complex projective space, expressing it in terms of the rank of the bundle and its degree. We give here a generalization to the $C_2$-equivariant context, using the calculation of the cohomology of a $C_2$-complex projective space from an earlier paper. We use ordinary $C_2$-coh… ▽ More Bézout's theorem, nonequivariantly, can be interpreted as a calculation of the Euler class of a sum of line bundles over complex projective space, expressing it in terms of the rank of the bundle and its degree. We give here a generalization to the $C_2$-equivariant context, using the calculation of the cohomology of a $C_2$-complex projective space from an earlier paper. We use ordinary $C_2$-cohomology with Burnside ring coefficients and an extended grading necessary to define the Euler class, which we express in terms of the equivariant rank of the bundle and the degrees of the bundle and its fixed subbundles. We do similar calculations using constant $\mathbb{Z}$ coefficients and Borel cohomology and compare the results. △ Less

Submitted 10 November, 2022; originally announced November 2022.

Comments: 20 pages

MSC Class: 55N91 (Primary); 14N10; 14N15; 55R40; 55R91 (Secondary)

A Bayesian constitutive model selection framework for biaxial mechanical testing of planar soft tissues: application to porcine aortic valves

Authors: Ankush Aggarwal, Luke T. Hudson, Devin W. Laurence, Chung-Hao Lee, Sanjay Pant

Abstract: A variety of constitutive models have been developed for soft tissue mechanics. However, there is no established criterion to select a suitable model for a specific application. Although the model that best fits the experimental data can be deemed the most suitable model, this practice often can be insufficient given the inter-sample variability of experimental observations. Herein, we present a B… ▽ More A variety of constitutive models have been developed for soft tissue mechanics. However, there is no established criterion to select a suitable model for a specific application. Although the model that best fits the experimental data can be deemed the most suitable model, this practice often can be insufficient given the inter-sample variability of experimental observations. Herein, we present a Bayesian approach to calculate the relative probabilities of constitutive models based on biaxial mechanical testing of tissue samples. 46 samples of porcine aortic valve tissue were tested using a biaxial stretching setup. For each sample, seven ratios of stresses along and perpendicular to the fiber direction were applied. The probabilities of eight invariant-based constitutive models were calculated based on the experimental data using the proposed model selection framework. The calculated probabilities showed that, out of the considered models and based on the information available through the utilized experimental dataset, the May--Newman model was the most probable model for the porcine aortic valve data. When the samples were grouped into different cusp types, the May--Newman model remained the most probable for the left- and right-coronary cusps, whereas for non-coronary cusps two models were found to be equally probable: the Lee--Sacks model and the May--Newman model. This difference between cusp types was found to be associated with the first principal component analysis (PCA) mode, where this mode's amplitudes of the non-coronary and right-coronary cusps were found to be significantly different. Our results show that a PCA-based statistical model can capture significant variations in the mechanical properties of soft tissues. The presented framework is applicable to any tissue type, and has the potential to provide a structured and rational way of making simulations population-based. △ Less

Submitted 3 January, 2023; v1 submitted 26 September, 2022; originally announced September 2022.

UNav: An Infrastructure-Independent Vision-Based Navigation System for People with Blindness and Low vision

Authors: Anbang Yang, Mahya Beheshti, Todd E Hudson, Rajesh Vedanthan, Wachara Riewpaiboon, Pattanasak Mongkolwat, Chen Feng, John-Ross Rizzo

Abstract: Vision-based localization approaches now underpin newly emerging navigation pipelines for myriad use cases from robotics to assistive technologies. Compared to sensor-based solutions, vision-based localization does not require pre-installed sensor infrastructure, which is costly, time-consuming, and/or often infeasible at scale. Herein, we propose a novel vision-based localization pipeline for a s… ▽ More Vision-based localization approaches now underpin newly emerging navigation pipelines for myriad use cases from robotics to assistive technologies. Compared to sensor-based solutions, vision-based localization does not require pre-installed sensor infrastructure, which is costly, time-consuming, and/or often infeasible at scale. Herein, we propose a novel vision-based localization pipeline for a specific use case: navigation support for end-users with blindness and low vision. Given a query image taken by an end-user on a mobile application, the pipeline leverages a visual place recognition (VPR) algorithm to find similar images in a reference image database of the target space. The geolocations of these similar images are utilized in downstream tasks that employ a weighted-average method to estimate the end-user's location and a perspective-n-point (PnP) algorithm to estimate the end-user's direction. Additionally, this system implements Dijkstra's algorithm to calculate a shortest path based on a navigable map that includes trip origin and destination. The topometric map used for localization and navigation is built using a customized graphical user interface that projects a 3D reconstructed sparse map, built from a sequence of images, to the corresponding a priori 2D floor plan. Sequential images used for map construction can be collected in a pre-map** step or scavenged through public databases/citizen science. The end-to-end system can be installed on any internet-accessible device with a camera that hosts a custom mobile application. For evaluation purposes, map** and localization were tested in a complex hospital environment. The evaluation results demonstrate that our system can achieve localization with an average error of less than 1 meter without knowledge of the camera's intrinsic parameters, such as focal length. △ Less

Submitted 22 September, 2022; originally announced September 2022.

MuLD: The Multitask Long Document Benchmark

Authors: G Thomas Hudson, Noura Al Moubayed

Abstract: The impressive progress in NLP techniques has been driven by the development of multi-task benchmarks such as GLUE and SuperGLUE. While these benchmarks focus on tasks for one or two input sentences, there has been exciting work in designing efficient techniques for processing much longer inputs. In this paper, we present MuLD: a new long document benchmark consisting of only documents over 10,000… ▽ More The impressive progress in NLP techniques has been driven by the development of multi-task benchmarks such as GLUE and SuperGLUE. While these benchmarks focus on tasks for one or two input sentences, there has been exciting work in designing efficient techniques for processing much longer inputs. In this paper, we present MuLD: a new long document benchmark consisting of only documents over 10,000 tokens. By modifying existing NLP tasks, we create a diverse benchmark which requires models to successfully model long-term dependencies in the text. We evaluate how existing models perform, and find that our benchmark is much more challenging than their `short document' equivalents. Furthermore, by evaluating both regular and efficient transformers, we show that models with increased context length are better able to solve the tasks presented, suggesting that future improvements in these models are vital for solving similar long document problems. We release the data and code for baselines to encourage further research on efficient NLP models. △ Less

Submitted 15 February, 2022; originally announced February 2022.

Network-Aware 5G Edge Computing for Object Detection: Augmenting Wearables to "See" More, Farther and Faster

Authors: Zhongzheng Yuan, Tommy Azzino, Yu Hao, Yixuan Lyu, Haoyang Pei, Alain Boldini, Marco Mezzavilla, Mahya Beheshti, Maurizio Porfiri, Todd Hudson, William Seiple, Yi Fang, Sundeep Rangan, Yao Wang, J. R. Rizzo

Abstract: Advanced wearable devices are increasingly incorporating high-resolution multi-camera systems. As state-of-the-art neural networks for processing the resulting image data are computationally demanding, there has been growing interest in leveraging fifth generation (5G) wireless connectivity and mobile edge computing for offloading this processing to the cloud. To assess this possibility, this pape… ▽ More Advanced wearable devices are increasingly incorporating high-resolution multi-camera systems. As state-of-the-art neural networks for processing the resulting image data are computationally demanding, there has been growing interest in leveraging fifth generation (5G) wireless connectivity and mobile edge computing for offloading this processing to the cloud. To assess this possibility, this paper presents a detailed simulation and evaluation of 5G wireless offloading for object detection within a powerful, new smart wearable called VIS4ION, for the Blind-and-Visually Impaired (BVI). The current VIS4ION system is an instrumented book-bag with high-resolution cameras, vision processing and haptic and audio feedback. The paper considers uploading the camera data to a mobile edge cloud to perform real-time object detection and transmitting the detection results back to the wearable. To determine the video requirements, the paper evaluates the impact of video bit rate and resolution on object detection accuracy and range. A new street scene dataset with labeled objects relevant to BVI navigation is leveraged for analysis. The vision evaluation is combined with a detailed full-stack wireless network simulation to determine the distribution of throughputs and delays with real navigation paths and ray-tracing from new high-resolution 3D models in an urban environment. For comparison, the wireless simulation considers both a standard 4G-Long Term Evolution (LTE) carrier and high-rate 5G millimeter-wave (mmWave) carrier. The work thus provides a thorough and realistic assessment of edge computing with mmWave connectivity in an application with both high bandwidth and low latency requirements. △ Less

Submitted 15 April, 2022; v1 submitted 25 December, 2021; originally announced December 2021.

Comments: Published in: IEEE Access ( Volume: 10)

The InSight HP$^3$ Penetrator (Mole) on Mars: Soil Properties Derived From the Penetration Attempts and Related Activities

Authors: T. Spohn, T. L. Hudson, E. Marteau, M. Golombek, M. Grott, T. Wippermann, K. S. Ali, C. Schmelzbach, S. Kedar, K. Hurst, A. Trebi-Ollennu, V. Ansan, J. Garvin, J. Knollenberg, N. Mueller, S. Piqeux, R. Lichtenheldt, C. Krause, C. Fantinati, N. Brinkman, D. Sollberger, P. Delage, C. Vrettos, S. Reershemius, L. Wisniewski , et al. (9 additional authors not shown)

Abstract: The NASA InSight Lander on Mars includes the Heat Flow and Physical Properties Package HP$^3$ to measure the surface heat flow of the planet. The package uses temperature sensors that would have been brought to the target depth of 3--5 m by a small penetrator, nicknamed the mole. The mole requiring friction on its hull to balance remaining recoil from its hammer mechanism did not penetrate to the… ▽ More The NASA InSight Lander on Mars includes the Heat Flow and Physical Properties Package HP$^3$ to measure the surface heat flow of the planet. The package uses temperature sensors that would have been brought to the target depth of 3--5 m by a small penetrator, nicknamed the mole. The mole requiring friction on its hull to balance remaining recoil from its hammer mechanism did not penetrate to the targeted depth. Instead, by precessing about a point midway along its hull, it carved a 7 cm deep and 5-6 cm wide pit and reached a depth of initially 31 cm. The root cause of the failure - as was determined through an extensive, almost two years long campaign - was a lack of friction in an unexpectedly thick cohesive duricrust. During the campaign -- described in detail in this paper -- the mole penetrated further aided by friction applied using the scoop at the end of the robotic Instrument Deployment Arm and by direct support by the latter. The mole finally reached a depth of 40 cm, bringing the mole body 1--2 cm below the surface. The penetration record of the mole and its thermal sensors were used to measure thermal and mechanical soil parameters such as the thermal conductivity and the penetration resistance of the duricrust and its cohesion. The hammerings of the mole were recorded by the seismometer SEIS and the signals could be used to derive a P-wave velocity and a S-wave velocity and elastic moduli representative of the topmost tens of cm of the regolith. The combined data were used to derive a model of the regolith that has an about 20 cm thick duricrust underneath a 1 cm thick unconsolidated layer of sand mixed with dust and above another 10 cm of unconsolidated sand. Underneath the latter, a layer more resistant to penetration and possibly consisting of debris from a small impact crater is inferred. △ Less

Submitted 8 December, 2021; originally announced December 2021.

Comments: 78 pages 22 figures, , submitted to Space Science Reviews

Report number: InSight contribution # 223

Journal ref: Space Science Reviews (2022) 218: 72

The InSight HP$^3$ mole on Mars: Lessons learned from attempts to penetrate to depth in the Martian soil

Authors: T. Spohn, T. L. Hudson, L. Witte, T. Wippermann, L. Wisniewski, B. Kediziora, C. Vrettos, R. D. Lorenz, M. Golombek, R. Lichtenfeld, M. Grott, J. Knollenberg, C. Krause, C. Fantinati, S. Nagihara, J. Grygorczuk

Abstract: The NASA InSight mission payload includes the Heat Flow and Physical Properties Package HP$^3$ to measure the surface heat flow. The package was designed to use a small penetrator -- nicknamed the mole -- to implement a string of temperature sensors in the soil to a depth of 5m. The mole itself is equipped with sensors to measure a thermal conductivity as it proceeds to depth. The heat flow would… ▽ More The NASA InSight mission payload includes the Heat Flow and Physical Properties Package HP$^3$ to measure the surface heat flow. The package was designed to use a small penetrator -- nicknamed the mole -- to implement a string of temperature sensors in the soil to a depth of 5m. The mole itself is equipped with sensors to measure a thermal conductivity as it proceeds to depth. The heat flow would be calculated from the product of the temperature gradient and the thermal conductivity. To avoid the perturbation caused by annual surface temperature variations, the measurements would be taken at a depth between 3 m and 5 m. The mole was designed to penetrate cohesionless soil similar to Quartz sand which was expected to provide a good analogue material for Martian sand. The sand would provide friction to the buried mole hull to balance the remaining recoil of the mole hammer mechanism that drives the mole forward. Unfortunately, the mole did not penetrate more than a mole length of 40 cm. The failure to penetrate deeper was largely due to a few tens of centimeter thick cohesive duricrust that failed to provide the required friction. Although a suppressor mass and spring in the hammer mechanism absorbed much of the recoil, the available mass did not allow a system that would have eliminated the recoil. The mole penetrated to 40 cm depth benefiting from friction provided by springs in the support structure from which it was deployed. It was found in addition that the Martian soil provided unexpected levels of penetration resistance that would have motivated to designing a more powerful mole. It is concluded that more mass would have allowed to design a more robust system with little or no recoil, more energy of the mole hammer mechanism and a more massive support structure. △ Less

Submitted 6 December, 2021; originally announced December 2021.

Comments: 34 pages, 15 figures, submitted to Adnaves in Space Research

Report number: InSight Contribution Number 234

Journal ref: Advances in Space Research, 69 (2022) 3140-3163

Elasto-plastic evolution of single crystals driven by dislocation flow

Authors: Thomas Hudson, Filip Rindler

Abstract: This work introduces a model for large-strain, geometrically nonlinear elasto-plastic dynamics in single crystals. The key feature of our model is that the plastic dynamics are entirely driven by the movement of dislocations, that is, $1$-dimensional topological defects in the crystal lattice. It is well known that glide motion of dislocations is the dominant microscopic mechanism for plastic defo… ▽ More This work introduces a model for large-strain, geometrically nonlinear elasto-plastic dynamics in single crystals. The key feature of our model is that the plastic dynamics are entirely driven by the movement of dislocations, that is, $1$-dimensional topological defects in the crystal lattice. It is well known that glide motion of dislocations is the dominant microscopic mechanism for plastic deformation in many crystalline materials, most notably in metals. We propose a novel geometric language, built on the concepts of space-time "slip trajectories" and the "crystal scaffold" to describe the movement of (discrete) dislocations and to couple this movement to plastic flow. The energetics and dissipation relationships in our model are derived from first principles drawing on the theories of crystal modeling, elasticity, and thermodynamics. The resulting force balances involve a new configurational stress tensor describing the forces acting against slip. In order to place our model into context, we further show that it recovers several laws that were known in special cases before, most notably the equation for the Peach-Koehler force (linearized configurational force) and the fact that the combination of all dislocations yields the curl of the plastic distortion field. Finally, we also include a brief discussion on how a number of other effects, such as hardening, softening, dislocation climb, and coarse-graining, could be incorporated into our model. △ Less

Submitted 9 February, 2022; v1 submitted 17 September, 2021; originally announced September 2021.

Comments: 44 pages, 5 figures

Asymptotic Expansion of the Elastic Far-Field of a Crystalline Defect

Authors: Julian Braun, Thomas Hudson, Christoph Ortner

Abstract: Lattice defects in crystalline materials create long-range elastic fields which can be modelled on the atomistic scale using an infinite system of discrete nonlinear force balance equations. Starting with these equations, this work rigorously derives a novel far-field expansion of these fields: The expansion is computable and is expressed as a sum of continuum correctors and discrete multipole ter… ▽ More Lattice defects in crystalline materials create long-range elastic fields which can be modelled on the atomistic scale using an infinite system of discrete nonlinear force balance equations. Starting with these equations, this work rigorously derives a novel far-field expansion of these fields: The expansion is computable and is expressed as a sum of continuum correctors and discrete multipole terms which decay with increasing algebraic rate as the order of the expansion increases. Truncating the expansion leaves a remainder describing the defect core structure, which is localised in the sense that it decays with an algebraic rate corresponding to the order at which the truncation occurred. △ Less

Submitted 10 August, 2021; originally announced August 2021.

MSC Class: 74G10; 70C20; 74E15; 41A58; 74G15

Recursive sequences attached to modular representations of finite groups

Authors: Alexandru Chirvasitu, Tara Hudson, Aparna Upadhyay

Abstract: The core of a finite-dimensional modular representation $M$ of a finite group $G$ is its largest non-projective summand. We prove that the dimensions of the cores of $M^{\otimes n}$ have algebraic Hilbert series when $M$ is Omega-algebraic, in the sense that the non-projective summands of $M^{\otimes n}$ fall into finitely many orbits under the action of the syzygy operator $Ω$. Similarly, we prov… ▽ More The core of a finite-dimensional modular representation $M$ of a finite group $G$ is its largest non-projective summand. We prove that the dimensions of the cores of $M^{\otimes n}$ have algebraic Hilbert series when $M$ is Omega-algebraic, in the sense that the non-projective summands of $M^{\otimes n}$ fall into finitely many orbits under the action of the syzygy operator $Ω$. Similarly, we prove that these dimension sequences are eventually linearly recursive when $M$ is what we term $Ω^{+}$-algebraic. This partially answers a conjecture by Benson and Symonds. Along the way, we also prove a number of auxiliary permanence results for linear recurrence under operations on multi-variable sequences. △ Less

Submitted 10 May, 2021; originally announced May 2021.

Comments: 30 pages + references

MSC Class: 20C05; 16D40; 13F25; 11K31

Witt groups of spinor varieties

Authors: Thomas Hudson, Arthur Martirosian, Heng Xie

Abstract: We show that Witt groups of spinor varieties (aka.\ maximal isotropic Grassmannians) can be presented by combinatorial objects called even shifted young diagram. Our method relies on the Blow-up setup of Balmer-Calmès, and we investigate the connecting homomorphism of the localization sequence via the projective bundle formula of Walter-Nenashev, the projection formula of Calmès-Hornbostel and the… ▽ More We show that Witt groups of spinor varieties (aka.\ maximal isotropic Grassmannians) can be presented by combinatorial objects called even shifted young diagram. Our method relies on the Blow-up setup of Balmer-Calmès, and we investigate the connecting homomorphism of the localization sequence via the projective bundle formula of Walter-Nenashev, the projection formula of Calmès-Hornbostel and the excess intersection formula of Fasel. △ Less

Submitted 5 July, 2022; v1 submitted 4 March, 2021; originally announced March 2021.

Comments: An appendix by Heng Xie; Final version. Generalized to the case of general bundles with complete flags

MSC Class: 11E81; 14M15; 19G99; and 14C20

Absolute energies and emission line shapes of the L x-ray transitions of lanthanide metals

Authors: Joseph W. Fowler, Galen C. O'Neil, Bradley K. Alpert, Douglas A. Bennett, Ed V. Denison, W. B. Doriese, Gene C. Hilton, Lawrence T. Hudson, Young-Il Joe, Kelsey M. Morgan, Daniel R. Schmidt, Daniel S. Swetz, Csilla I. Szabo, Joel N. Ullom

Abstract: We use an array of transition-edge sensors, cryogenic microcalorimeters with 4 eV energy resolution, to measure L x-ray emission-line profiles of four elements of the lanthanide series: praseodymium, neodymium, terbium, and holmium. The spectrometer also surveys numerous x-ray standards in order to establish an absolute-energy calibration traceable to the International System of Units for the ener… ▽ More We use an array of transition-edge sensors, cryogenic microcalorimeters with 4 eV energy resolution, to measure L x-ray emission-line profiles of four elements of the lanthanide series: praseodymium, neodymium, terbium, and holmium. The spectrometer also surveys numerous x-ray standards in order to establish an absolute-energy calibration traceable to the International System of Units for the energy range 4 keV to 10 keV. The new results include emission line profiles for 97 lines, each expressed as a sum of one or more Voigt functions; improved absolute energy uncertainty on 71 of these lines relative to existing reference data; a median uncertainty on the peak energy of 0.24 eV, four to ten times better than the median of prior work; and 6 lines that lack any measured values in existing reference tables. The 97 lines comprise nearly all of the most intense L lines from these elements under broad-band x-ray excitation. The work improves on previous measurements made with a similar cryogenic spectrometer by the use of sensors with better linearity in the absorbed energy and a gold x-ray absorbing layer that has a Gaussian energy-response function. It also employs a novel sample holder that enables rapid switching between science targets and calibration targets with excellent gain balancing. Most of the results for peak energy values shown here should be considered as replacements for the currently tabulated standard reference values, while the line shapes given here represent a significant expansion of the scope of available reference data. △ Less

Submitted 30 November, 2020; originally announced December 2020.

Journal ref: Metrologia 58 (2021) 015016

A Systematic Analysis of the Memory term in Coarse-Grained models: the case of the Markovian Approximation

Authors: N. Di Pasquale, T. Hudson, M. Icardi, L. Rovigatti, M. Spinaci

Abstract: The systematic development of Coarse-Grained (CG) models via the Mori-Zwanzig projector operator formalism requires the explicit description of several terms, including a deterministic drift term, a dissipative memory term and a random fluctuation term. In many applications, the memory and fluctuation terms are related by the fluctuation-dissipation relation and are, in general, more challenging t… ▽ More The systematic development of Coarse-Grained (CG) models via the Mori-Zwanzig projector operator formalism requires the explicit description of several terms, including a deterministic drift term, a dissipative memory term and a random fluctuation term. In many applications, the memory and fluctuation terms are related by the fluctuation-dissipation relation and are, in general, more challenging to derive than the drift term. In this work we analyse an approximation of the memory term and propose a rational basis for a data-driven approach to an approximation of the memory and fluctuating terms which can be considered included in the class of the Markovian ones. △ Less

Submitted 28 September, 2021; v1 submitted 30 October, 2020; originally announced November 2020.

Comments: 39 pages, 2 Figures

Learning to Play No-Press Diplomacy with Best Response Policy Iteration

Authors: Thomas Anthony, Tom Eccles, Andrea Tacchetti, János Kramár, Ian Gemp, Thomas C. Hudson, Nicolas Porcel, Marc Lanctot, Julien Pérolat, Richard Everett, Roman Werpachowski, Satinder Singh, Thore Graepel, Yoram Bachrach

Abstract: Recent advances in deep reinforcement learning (RL) have led to considerable progress in many 2-player zero-sum games, such as Go, Poker and Starcraft. The purely adversarial nature of such games allows for conceptually simple and principled application of RL methods. However real-world settings are many-agent, and agent interactions are complex mixtures of common-interest and competitive aspects.… ▽ More Recent advances in deep reinforcement learning (RL) have led to considerable progress in many 2-player zero-sum games, such as Go, Poker and Starcraft. The purely adversarial nature of such games allows for conceptually simple and principled application of RL methods. However real-world settings are many-agent, and agent interactions are complex mixtures of common-interest and competitive aspects. We consider Diplomacy, a 7-player board game designed to accentuate dilemmas resulting from many-agent interactions. It also features a large combinatorial action space and simultaneous moves, which are challenging for RL algorithms. We propose a simple yet effective approximate best response operator, designed to handle large combinatorial action spaces and simultaneous moves. We also introduce a family of policy iteration methods that approximate fictitious play. With these methods, we successfully apply RL to Diplomacy: we show that our agents convincingly outperform the previous state-of-the-art, and game theoretic equilibrium analysis shows that the new process yields consistent improvements. △ Less

Submitted 4 January, 2022; v1 submitted 8 June, 2020; originally announced June 2020.

Atomistic origins of continuum dislocation dynamics

Authors: Thomas Hudson, Patrick van Meurs, Mark A. Peletier

Abstract: This paper focuses on the connections between four stochastic and deterministic models for the motion of straight screw dislocations. Starting from a description of screw dislocation motion as interacting random walks on a lattice, we prove explicit estimates of the distance between solutions of this model, an SDE system for the dislocation positions, and two deterministic mean-field models descri… ▽ More This paper focuses on the connections between four stochastic and deterministic models for the motion of straight screw dislocations. Starting from a description of screw dislocation motion as interacting random walks on a lattice, we prove explicit estimates of the distance between solutions of this model, an SDE system for the dislocation positions, and two deterministic mean-field models describing the dislocation density. The proof of these estimates uses a collection of various techniques in analysis and probability theory, including a novel approach to establish propagation-of-chaos on a spatially discrete model. The estimates are non-asymptotic and explicit in terms of four parameters: the lattice spacing, the number of dislocations, the dislocation core size, and the temperature. This work is a first step in exploring this parameter space with the ultimate aim to connect and quantify the relationships between the many different dislocation models present in the literature. △ Less

Submitted 10 September, 2020; v1 submitted 16 January, 2020; originally announced January 2020.

Comments: Version 2, 63 pages

A Secure Cloud with Minimal Provider Trust

Authors: Amin Mosayyebzadeh, Gerardo Ravago, Apoorve Mohan, Ali Raza, Sahil Tikale, Nabil Schear, Trammell Hudson, Jason Hennessey, Naved Ansari, Kyle Hogan, Charles Munson, Larry Rudolph, Gene Cooperman, Peter Desnoyers, Orran Krieger

Abstract: Bolted is a new architecture for a bare metal cloud with the goal of providing security-sensitive customers of a cloud the same level of security and control that they can obtain in their own private data centers. It allows tenants to elastically allocate secure resources within a cloud while being protected from other previous, current, and future tenants of the cloud. The provisioning of a new s… ▽ More Bolted is a new architecture for a bare metal cloud with the goal of providing security-sensitive customers of a cloud the same level of security and control that they can obtain in their own private data centers. It allows tenants to elastically allocate secure resources within a cloud while being protected from other previous, current, and future tenants of the cloud. The provisioning of a new server to a tenant isolates a bare metal server, only allowing it to communicate with other tenant's servers once its critical firmware and software have been attested to the tenant. Tenants, rather than the provider, control the tradeoffs between security, price, and performance. A prototype demonstrates scalable end-to-end security with small overhead compared to a less secure alternative. △ Less

Submitted 13 July, 2019; originally announced July 2019.

Comments: 7 Pages, 10th USENIX Workshop on Hot Topics in Cloud Computing (HotCloud '18). arXiv admin note: text overlap with arXiv:1907.06110

Supporting Security Sensitive Tenants in a Bare-Metal Cloud

Authors: Amin Mosayyebzadeh, Apoorve Mohan, Sahil Tikale, Mania Abdi, Nabil Schear, Charles Munson, Trammell Hudson, Larry Rudolph, Gene Cooperman, Peter Desnoyers, Orran Krieger

Abstract: Bolted is a new architecture for bare-metal clouds that enables tenants to control tradeoffs between security, price, and performance. Security-sensitive tenants can minimize their trust in the public cloud provider and achieve similar levels of security and control that they can obtain in their own private data centers. At the same time, Bolted neither imposes overhead on tenants that are securit… ▽ More Bolted is a new architecture for bare-metal clouds that enables tenants to control tradeoffs between security, price, and performance. Security-sensitive tenants can minimize their trust in the public cloud provider and achieve similar levels of security and control that they can obtain in their own private data centers. At the same time, Bolted neither imposes overhead on tenants that are security insensitive nor compromises the flexibility or operational efficiency of the provider. Our prototype exploits a novel provisioning system and specialized firmware to enable elasticity similar to virtualized clouds. Experimentally we quantify the cost of different levels of security for a variety of workloads and demonstrate the value of giving control to the tenant. △ Less

Submitted 13 July, 2019; originally announced July 2019.

Comments: 16 Pages, 2019 USENIX Annual Technical Conference (ATC'19)

Stability of Bott--Samelson Classes in Algebraic Cobordism

Authors: Thomas Hudson, Tomoo Matsumura, Nicolas Perrin

Abstract: In this paper, we construct stable Bott--Samelson classes in the projective limit of the algebraic cobordism rings of full flag varieties, upon an initial choice of a reduced word in a given dimension. Each stable Bott--Samelson class is represented by a bounded formal power series modulo symmetric functions in positive degree. We make some explicit computations for those power series in the case… ▽ More In this paper, we construct stable Bott--Samelson classes in the projective limit of the algebraic cobordism rings of full flag varieties, upon an initial choice of a reduced word in a given dimension. Each stable Bott--Samelson class is represented by a bounded formal power series modulo symmetric functions in positive degree. We make some explicit computations for those power series in the case of infinitesimal cohomology. We also obtain a formula of the restriction of Bott--Samelson classes to smaller flag varieties. △ Less

Submitted 13 July, 2019; originally announced July 2019.

Comments: 23 pages. Comments are welcome

Report number: Springer Proceedings in Mathematics & Statistics, vol 332

Analysis of cell size effects in atomistic crack propagation

Authors: Maciej Buze, Thomas Hudson, Christoph Ortner

Abstract: We consider crack propagation in a crystalline material in terms of bifurcation analysis. We provide evidence that the stress intensity factor is a natural bifurcation parameter, and that the resulting bifurcation diagram is a periodic "snaking curve". We then prove qualitative properties of the equilibria and convergence rates of finite-cell approximations to the "exact" bifurcation diagram. We consider crack propagation in a crystalline material in terms of bifurcation analysis. We provide evidence that the stress intensity factor is a natural bifurcation parameter, and that the resulting bifurcation diagram is a periodic "snaking curve". We then prove qualitative properties of the equilibria and convergence rates of finite-cell approximations to the "exact" bifurcation diagram. △ Less

Submitted 26 February, 2020; v1 submitted 30 May, 2019; originally announced May 2019.

MSC Class: 65L20; 70C20; 74A45; 74G20; 74G40; 74G60; 74G65

On the K-theoretic fundamental classes of Deligne-Lusztig varieties

Authors: Thomas Hudson, Dennis Peters

Abstract: In this paper we express the class of the structure sheaves of the closures of Deligne--Lusztig varieties as explicit double Grothendieck polynomials in the first Chern classes of appropriate line bundles on the ambient flag variety. This is achieved by viewing such closures as degeneracy loci of morphisms of vector bundles. In this paper we express the class of the structure sheaves of the closures of Deligne--Lusztig varieties as explicit double Grothendieck polynomials in the first Chern classes of appropriate line bundles on the ambient flag variety. This is achieved by viewing such closures as degeneracy loci of morphisms of vector bundles. △ Less

Submitted 7 February, 2020; v1 submitted 17 April, 2019; originally announced April 2019.

Comments: 8 pages, minor changes

MSC Class: 20G40; 19E08 (Primary); 14M15 (Secondary)

Journal ref: Journal of Pure and Applied Algebra Volume 224, Issue 8, August 2020, 106335

Packing Concave Molecules in Crystals and Amorphous Solids: On the Connection between Shape and Local Structure

Authors: Cerridwen Jennings, Malcolm Ramsay, Toby Hudson, Peter Harrowell

Abstract: The structure of the densest crystal packings is determined for a variety of concave shapes in 2D constructed by the overlap of two or three disks. The maximum contact number per particle pair is defined and proposed as a useful means of categorizing particle shape. We demonstrate that the densest packed crystal exhibits a maximum in the number of contacts per particle but does not necessarily inc… ▽ More The structure of the densest crystal packings is determined for a variety of concave shapes in 2D constructed by the overlap of two or three disks. The maximum contact number per particle pair is defined and proposed as a useful means of categorizing particle shape. We demonstrate that the densest packed crystal exhibits a maximum in the number of contacts per particle but does not necessarily include particle pairs with the maximum contact number. In contrast, amorphous structures, generated by energy minimization of high temperature liquids, typically do include maximum contact pairs. The amorphous structures exhibit a large number of contacts per particle corresponding to over-constrained structures. Possible consequences of this over-constraint are discussed. △ Less

Submitted 7 February, 2019; originally announced February 2019.

Journal ref: Molecular Physics, 113 (17-18), 2755-2769 (2015)

The $C_2$-equivariant cohomology of complex projective spaces

Authors: Steven R. Costenoble, Thomas Hudson, Sean Tilson

Abstract: We compute the equivariant cohomology of complex projective spaces associated to finite-dimensional representations of $C_2$, using ordinary cohomology graded on representations of the fundamental groupoid, with coefficients in the Burnside ring Mackey functor. This extension of the $RO(C_2)$-graded theory allows for the definition of Euler classes, which are used as generators of the cohomology o… ▽ More We compute the equivariant cohomology of complex projective spaces associated to finite-dimensional representations of $C_2$, using ordinary cohomology graded on representations of the fundamental groupoid, with coefficients in the Burnside ring Mackey functor. This extension of the $RO(C_2)$-graded theory allows for the definition of Euler classes, which are used as generators of the cohomology of the projective spaces. As an application, we give an equivariant version of Bezout's theorem. △ Less

Submitted 21 June, 2021; v1 submitted 18 November, 2018; originally announced November 2018.

Comments: 57 pages, this is a major update. The proof of the main result has been simplified. As an application an equivariant version of Bezout's theorem was added

MSC Class: 55N91 (Primary) 14N10; 14N15; 55R40; 55R91 (Secondary)

Journal ref: Advances in Mathematics Volume 398, 26 March 2022, 108245

Coarse-graining of overdamped Langevin dynamics via the Mori-Zwanzig formalism

Authors: Thomas Hudson, Xingjie Helen Li

Abstract: The Mori-Zwanzig formalism is applied to derive an equation for the evolution of linear observables of the overdamped Langevin equation. To illustrate the resulting equation and its use in deriving approximate models, a particular benchmark example is studied both numerically and via a formal asymptotic expansion. The example considered demonstrates the important of memory effects in determining t… ▽ More The Mori-Zwanzig formalism is applied to derive an equation for the evolution of linear observables of the overdamped Langevin equation. To illustrate the resulting equation and its use in deriving approximate models, a particular benchmark example is studied both numerically and via a formal asymptotic expansion. The example considered demonstrates the important of memory effects in determining the correct temporal behaviour of such systems. △ Less

Submitted 18 October, 2018; originally announced October 2018.

Comments: 19 pages, 5 figures

MSC Class: 41A60; 82C31; 60H10

Analysis of an atomistic model for anti-plane fracture

Authors: Maciej Buze, Thomas Hudson, Christoph Ortner

Abstract: We develop a model for an anti-plane crack defect posed on a square lattice under an interatomic pair-potential with nearest-neighbour interactions. In particular, we establish existence, local uniqueness and stability of solutions for small loading parameters and further prove qualitatively sharp far-field decay estimates. The latter requires establishing decay estimates for the corresponding lat… ▽ More We develop a model for an anti-plane crack defect posed on a square lattice under an interatomic pair-potential with nearest-neighbour interactions. In particular, we establish existence, local uniqueness and stability of solutions for small loading parameters and further prove qualitatively sharp far-field decay estimates. The latter requires establishing decay estimates for the corresponding lattice Green's function, which are of independent interest. △ Less

Submitted 6 November, 2019; v1 submitted 12 October, 2018; originally announced October 2018.

MSC Class: 65L20; 70C20; 74A45; 74G20; 74G40; 74G65

Double Grothendieck Polynomials for Symplectic and Odd Orthogonal Grassmannians

Authors: Thomas Hudson, Takeshi Ikeda, Tomoo Matsumura, Hiroshi Naruse

Abstract: We study the double Grothendieck polynomials of Kirillov--Naruse for the symplectic and odd orthogonal Grassmannians. These functions are explicitly written as sums of Pfaffian and are identified with the stable limits of the fundamental classes of Schubert varieties in the torus equivariant connective K-theory of these isotropic Grassmannians. We also provide a combinatorial description of the ri… ▽ More We study the double Grothendieck polynomials of Kirillov--Naruse for the symplectic and odd orthogonal Grassmannians. These functions are explicitly written as sums of Pfaffian and are identified with the stable limits of the fundamental classes of Schubert varieties in the torus equivariant connective K-theory of these isotropic Grassmannians. We also provide a combinatorial description of the ring formally spanned by double Grothendieck polynomials. △ Less

Submitted 20 September, 2018; originally announced September 2018.

Comments: This article was originally a part of our previous paper [arXiv:1504.02828]. It was separated from its published version and became an independent paper

Journal ref: J. Algebra, 546 (2020), pp. 294--314

An existence result for Discrete Dislocation Dynamics in three dimensions

Authors: Thomas Hudson

Abstract: We present a mathematical framework within which Discrete Dislocation Dynamics in three dimensions is well-posed. By considering smooth distributions of slip, we derive a regularised energy for curved dislocations, and rigorously derive the Peach-Koehler force on the dislocation network via an inner variation. We propose a dissipative evolution law which is cast as a generalised gradient flow, and… ▽ More We present a mathematical framework within which Discrete Dislocation Dynamics in three dimensions is well-posed. By considering smooth distributions of slip, we derive a regularised energy for curved dislocations, and rigorously derive the Peach-Koehler force on the dislocation network via an inner variation. We propose a dissipative evolution law which is cast as a generalised gradient flow, and using a discrete-in-time approximation scheme, existence and regularity results are obtained for the evolution, up until the first time at which an infinite density of dislocation lines forms. △ Less

Submitted 1 June, 2018; originally announced June 2018.

Comments: 30 pages, 3 figures

MSC Class: 35Q74; 74N05; 37N15

Systematic derivation of hybrid coarse-grained models

Authors: Nicodemo Di Pasquale, Thomas Hudson, Matteo Icardi

Abstract: Significant efforts have been devoted in the last decade towards improving the predictivity of coarse-grained models in molecular dynamics simulations and providing a rigorous justification of their use, through a combination of theoretical studies and data-driven approaches. One of the most promising research effort is the (re-)discovery of the Mori-Zwanzig projection as a generic, yet systematic… ▽ More Significant efforts have been devoted in the last decade towards improving the predictivity of coarse-grained models in molecular dynamics simulations and providing a rigorous justification of their use, through a combination of theoretical studies and data-driven approaches. One of the most promising research effort is the (re-)discovery of the Mori-Zwanzig projection as a generic, yet systematic, theoretical tool for deriving coarse-grained models. Despite its clean mathematical formulation and generality, there are still many open questions about its applicability and assumptions. In this work, we propose a detailed derivation of a hybrid multi-scale system, generalising and further investigating the approach developed in [Español, P., EPL, 88, 40008 (2009)]. Issues such as the general co-existence of atoms (fully-resolved degrees of freedom) and beads (larger coarse-grained units), the role of the fine-to-coarse map** chosen, and the approximation of effective potentials are discussed. The concept of an approximate projection is introduced along with a discussion of its use as measure of the error committed with the approximation of the true interactions among the beads. The theoretical discussion is supported by numerical simulations of a monodimensional non-linear periodic benchmark system with an open-source parallel Julia code, easily extensible to arbitrary potential models and fine-to-coarse map** functions. The results presented highlight the importance of introducing, in the macroscopic model, a non-constant dissipative term, given by the Mori-Zwanzig approach, to correctly reproduce the reference fine-grained results without requiring \emph{ad-hoc} calibration of interaction potentials and thermostats. △ Less

Submitted 6 August, 2018; v1 submitted 22 April, 2018; originally announced April 2018.

Comments: 29 pages, 9 figures

Journal ref: Phys. Rev. E 99, 013303 (2019)

Stochastic homogenization of a scalar viscoelastic model exhibiting stress-strain hysteresis

Authors: Thomas Hudson, Frédéric Legoll, Tony Lelièvre

Abstract: Motivated by rate-independent stress-strain hysteresis observed in filled rubber, this article considers a scalar viscoelastic model in which the constitutive law is random and varies on a lengthscale which is small relative to the overall size of the solid. Using stochastic two-scale convergence as introduced by Bourgeat, Mikelic and Wright, we obtain the homogenized limit of the evolution, and d… ▽ More Motivated by rate-independent stress-strain hysteresis observed in filled rubber, this article considers a scalar viscoelastic model in which the constitutive law is random and varies on a lengthscale which is small relative to the overall size of the solid. Using stochastic two-scale convergence as introduced by Bourgeat, Mikelic and Wright, we obtain the homogenized limit of the evolution, and demonstrate that under certain hypotheses, the homogenized model exhibits hysteretic behaviour which persists under asymptotically slow loading. These results are illustrated by means of numerical simulations in a particular one-dimensional instance of the model. △ Less

Submitted 21 November, 2019; v1 submitted 15 February, 2018; originally announced February 2018.

Comments: 54 pages, 4 figures

MSC Class: 74Q10; 35Q74

Symplectic and odd orthogonal Pfaffian formulas for algebraic cobordism

Authors: Thomas Hudson, Tomoo Matsumura

Abstract: In this paper we provide generalisations of Pfaffian formulas for the degeneracy loci classes in the algebraic cobordism of symplectic/odd orthogonal Grassmann bundles. In this paper we provide generalisations of Pfaffian formulas for the degeneracy loci classes in the algebraic cobordism of symplectic/odd orthogonal Grassmann bundles. △ Less

Submitted 19 October, 2017; originally announced October 2017.

Comments: 19 pages

MSC Class: (Primary) 14M15; 55N22; (Secondary) 05E05; 14C17;

Journal ref: Pacific J. Math. 302 (2019) 97-118

Properties of screw dislocation dynamics: time estimates on boundary and interior collisions

Authors: Thomas Hudson, Marco Morandotti

Abstract: In this paper, the dynamics of a system of a finite number of screw dislocations is studied. Under the assumption of antiplane linear elasticity, the two-dimensional dynamics is determined by the renormalised energy. The interaction of one dislocation with the boundary and of two dislocations of opposite Burgers moduli are analysed in detail and estimates on the collision times are obtained. Some… ▽ More In this paper, the dynamics of a system of a finite number of screw dislocations is studied. Under the assumption of antiplane linear elasticity, the two-dimensional dynamics is determined by the renormalised energy. The interaction of one dislocation with the boundary and of two dislocations of opposite Burgers moduli are analysed in detail and estimates on the collision times are obtained. Some exactly solvable cases and numerical simulations show agreement with the estimates obtained. △ Less

Submitted 29 June, 2017; v1 submitted 7 March, 2017; originally announced March 2017.

Comments: 25 pages, 4 figures

MSC Class: 70Fxx; 37N15; 74H05; 82D25

A Reassessment of Absolute Energies of the X-ray L Lines of Lanthanide Metals

Authors: J. W. Fowler, B. K. Alpert, D. A. Bennett, W. B. Doriese, J. D. Gard, G. C. Hilton, L. T. Hudson, Y. -I. Joe, K. M. Morgan, G. C. O'Neil, C. D. Reintsema, D. R. Schmidt, D. S. Swetz, C. I. Szabo, J. N. Ullom.

Abstract: We introduce a new technique for determining x-ray fluorescence line energies and widths, and we present measurements made with this technique of 22 x-ray L lines from lanthanide-series elements. The technique uses arrays of transition-edge sensors, microcalorimeters with high energy-resolving power that simultaneously observe both calibrated x-ray standards and the x-ray emission lines under stud… ▽ More We introduce a new technique for determining x-ray fluorescence line energies and widths, and we present measurements made with this technique of 22 x-ray L lines from lanthanide-series elements. The technique uses arrays of transition-edge sensors, microcalorimeters with high energy-resolving power that simultaneously observe both calibrated x-ray standards and the x-ray emission lines under study. The uncertainty in absolute line energies is generally less than 0.4 eV in the energy range of 4.5 keV to 7.5 keV. Of the seventeen line energies of neodymium, samarium, and holmium, thirteen are found to be consistent with the available x-ray reference data measured after 1990; only two of the four lines for which reference data predate 1980, however, are consistent with our results. Five lines of terbium are measured with uncertainties that improve on those of existing data by factors of two or more. These results eliminate a significant discrepancy between measured and calculated x-ray line energies for the terbium Ll line (5.551 keV). The line widths are also measured, with uncertainties of 0.6 eV or less on the full-width at half-maximum in most cases. These measurements were made with an array of approximately one hundred superconducting x- ray microcalorimeters, each sensitive to an energy band from 1 keV to 8 keV. No energy-dispersive spectrometer has previously been used for absolute-energy estimation at this level of accuracy. Future spectrometers, with superior linearity and energy resolution, will allow us to improve on these results and expand the measurements to more elements and a wider range of line energies. △ Less

Submitted 10 May, 2017; v1 submitted 1 February, 2017; originally announced February 2017.

Comments: This version is the one accepted for publication in Metrologia

Journal ref: Metrologia 54 (2017) 494

Vexillary degeneracy loci classes in K-theory and algebraic cobordism

Authors: Thomas Hudson, Tomoo Matsumura

Abstract: In this paper, we prove determinant formulas for the $K$-theory classes of the structure sheaves of degeneracy loci classes associated to vexillary permutations in type $A$. As a consequence we obtain determinant formulas for Lascoux-Schützenberger's double Grothendieck polynomials associated to vexillary permutations. Furthermore, we generalize the determinant formula to algebraic cobordism. In this paper, we prove determinant formulas for the $K$-theory classes of the structure sheaves of degeneracy loci classes associated to vexillary permutations in type $A$. As a consequence we obtain determinant formulas for Lascoux-Schützenberger's double Grothendieck polynomials associated to vexillary permutations. Furthermore, we generalize the determinant formula to algebraic cobordism. △ Less

Submitted 1 January, 2017; originally announced January 2017.

Comments: 12 pages

Phase diagram of heteronuclear Janus dumbbells

Authors: Patrick O'Toole, Achille Giacometti, Toby Hudson

Abstract: Using Aggregation-Volume-Bias Monte Carlo simulations along with Successive Umbrella Sampling and Histogram Re-weighting, we study the phase diagram of a system of dumbbells formed by two touching spheres having variable sizes, as well as different interaction properties. The first sphere ($h$) interacts with all other spheres belonging to different dumbbells with a hard-sphere potential. The seco… ▽ More Using Aggregation-Volume-Bias Monte Carlo simulations along with Successive Umbrella Sampling and Histogram Re-weighting, we study the phase diagram of a system of dumbbells formed by two touching spheres having variable sizes, as well as different interaction properties. The first sphere ($h$) interacts with all other spheres belonging to different dumbbells with a hard-sphere potential. The second sphere ($s$) interacts via a square-well interaction with other $s$ spheres belonging to different dumbbells and with a hard-sphere potential with all remaining $h$ spheres. We focus on the region where the $s$ sphere is larger than the $h$ sphere, as measured by a parameter $1\le α\le 2 $ controlling the relative size of the two spheres. As $α\to 2$ a simple fluid of square-well spheres is recovered, whereas $α\to 1$ corresponds to the Janus dumbbell limit, where the $h$ and $s$ spheres have equal sizes. Many phase diagrams falling into three classes are observed, depending on the value of $α$. The $1.8 \le α\le 2$ is dominated by a gas-liquid phase separation very similar to that of a pure square-well fluid with varied critical temperature and density. When $1.3 \le α\le 1.8$ we find a progressive destabilization of the gas-liquid phase diagram by the onset of self-assembled structures, that eventually lead to a metastability of the gas-liquid transition below $α=1.2$. △ Less

Submitted 18 December, 2016; originally announced December 2016.

Comments: 16 pages, 14 Figures, accepted for publication in Soft Matter

Asymptotic analysis of boundary layers in a repulsive particle system

Authors: Cameron Hall, Thomas Hudson, Patrick van Meurs

Abstract: This paper studies the boundary behaviour at mechanical equilibrium at the ends of a finite interval of a class of systems of interacting particles with monotone decreasing repulsive force. Our setting covers pile-ups of dislocations, dislocation dipoles and dislocation walls. The main challenge is to control the nonlocal nature of the pairwise particle interactions. Using matched asymptotic expan… ▽ More This paper studies the boundary behaviour at mechanical equilibrium at the ends of a finite interval of a class of systems of interacting particles with monotone decreasing repulsive force. Our setting covers pile-ups of dislocations, dislocation dipoles and dislocation walls. The main challenge is to control the nonlocal nature of the pairwise particle interactions. Using matched asymptotic expansions for the particle positions and rigorous development of an appropriate energy via Gamma-convergence, we obtain the equilibrium equation solved by the boundary layer correction, associate an energy with an appropriate scaling to this correction, and provide decay rates into the bulk. △ Less

Submitted 11 September, 2016; originally announced September 2016.

Comments: 41 pages, 13 figures

MSC Class: 74Q05; 74G10; 41A60

Kempf-Laksov Schubert classes for even infinitesimal cohomology theories

Authors: Thomas Hudson, Tomoo Matsumura

Abstract: In this paper, we prove a generalization of Kempf-Laksov formula for the degeneracy loci classes in even infinitesimal cohomology theories of the Grassmannian bundle and the Lagrangian Grassmannian bundle. In this paper, we prove a generalization of Kempf-Laksov formula for the degeneracy loci classes in even infinitesimal cohomology theories of the Grassmannian bundle and the Lagrangian Grassmannian bundle. △ Less

Submitted 25 February, 2016; originally announced February 2016.

Comments: 22 pages. arXiv admin note: text overlap with arXiv:1602.05704

Journal ref: EMS Series of Congress Reports- Schubert Varieties, Equivariant Cohomology and Characteristic Classes- Impanga 15, 2018. Pages 127-151

Segre classes and Kempf-Laksov formula in algebraic cobordism

Authors: Thomas Hudson, Tomoo Matsumura

Abstract: In this paper, we study Segre classes in algebraic cobordism. We also prove a generalization of Kempf-Laksov formula for the degeneracy loci classes in the algebraic cobordism of the Grassmannian bundle. In this paper, we study Segre classes in algebraic cobordism. We also prove a generalization of Kempf-Laksov formula for the degeneracy loci classes in the algebraic cobordism of the Grassmannian bundle. △ Less

Submitted 25 April, 2018; v1 submitted 18 February, 2016; originally announced February 2016.

Comments: 14 pages. The exposition and some arguments have been modified

MSC Class: 14M15; 55N22; 05E05; 14C17

Journal ref: Mathematische Annalen, 2019, Volume 374, Issue 3-4, pp 1439-1457

Density and Glass Forming Ability in Amorphous Atomic Alloys: the Role of the Particle Softness

Authors: Ian Douglass, Toby Hudson, Peter Harrowell

Abstract: A key property of glass forming alloys, the anomalously small volume difference with respect to the crystal, is shown to arise as a direct consequence of the soft repulsive potentials between metals. This feature of the inter-atomic potential is demonstrated to be responsible for a significant component of the glass forming ability of alloys due to the decrease in the enthalpy of fusion and the as… ▽ More A key property of glass forming alloys, the anomalously small volume difference with respect to the crystal, is shown to arise as a direct consequence of the soft repulsive potentials between metals. This feature of the inter-atomic potential is demonstrated to be responsible for a significant component of the glass forming ability of alloys due to the decrease in the enthalpy of fusion and the associated depression of the freezing point. △ Less

Submitted 16 February, 2016; originally announced February 2016.

Long Range Stress Correlations in the Inherent Structures of Liquids at Rest

Authors: Sadrul Chowdhury, Sneha Abraham, Toby Hudson, Peter Harrowell

Abstract: Simulation studies of the atomic shear stress in the local potential energy minima (inherent structures) are reported for binary liquid mixtures in 2D and 3D. These inherent structure stresses are fundamental to slow stress relaxation and high viscosity in supercooled liquids. We find that the atomic shear stress in the inherent structures (IS) of both liquids at rest exhibits slowly decaying anis… ▽ More Simulation studies of the atomic shear stress in the local potential energy minima (inherent structures) are reported for binary liquid mixtures in 2D and 3D. These inherent structure stresses are fundamental to slow stress relaxation and high viscosity in supercooled liquids. We find that the atomic shear stress in the inherent structures (IS) of both liquids at rest exhibits slowly decaying anisotropic correlations. We show that the stress correlations contributes significantly to the variance of the total shear stress of the IS configurations and consider the origins of the anisotropy and spatial extent of the stress correlations. △ Less

Submitted 15 March, 2016; v1 submitted 16 February, 2016; originally announced February 2016.