Search

In this paper we argue that key, often sensational and misleading, claims regarding linguistic capabilities of Large Language Models (LLMs) are based on at least two unfounded assumptions; the assumption of language completeness and the assumption of data completeness. Language completeness assumes that a distinct and complete thing such as `a natural language' exists, the essential characteristics of which can be effectively and comprehensively modelled by an LLM. The assumption of data completeness relies on the belief that a language can be quantified and wholly captured by data. Work within the enactive approach to cognitive science makes clear that, rather than a distinct and complete thing, language is a means or way of acting. Languaging is not the kind of thing that can admit of a complete or comprehensive modelling. From an enactive perspective we identify three key characteristics of enacted language; embodiment, participation, and precariousness, that are absent in LLMs, and likely incompatible in principle with current architectures. We argue that these absences imply that LLMs are not now and cannot in their present form be linguistic agents the way humans are. We illustrate the point in particular through the phenomenon of `algospeak', a recently described pattern of high stakes human language activity in heavily controlled online environments. On the basis of these points, we conclude that sensational and misleading claims about LLM agency and capabilities emerge from a deep misconception of both what human language is and what LLMs are., Comment: To appear in the Journal of Language Sciences

We have derived the governing equations for an SIR model with delay terms in both the infectivity and recovery of the disease. The equations are derived by modelling the dynamics as a continuous time random walk, where individuals move between the classic SIR compartments. With an appropriate choice of distributions for the infectivity and recovery processes delay terms are introduced into the governing equations in a manner that ensures the physicality of the model. This provides novel insight into the underlying dynamics of an SIR model with time delays. The SIR model with delay infectivity and recovery allows for a more diverse range of dynamical behaviours. The model accounts for an incubation effect without the need to introduce new compartments., Comment: 17 pages, 6 figures

Compartment models with delay terms are widely used across a range of disciplines. The motivation to include delay terms varies across different contexts. In epidemiological and pharmacokinetic models, the delays are often used to represent an incubation period. In this work, we derive a compartment model with delay terms from an underlying non-Markov stochastic process. Delay terms arise when waiting times are drawn from a delay exponential distribution. This stochastic process approach allows us to preserve the physicality of the model, gaining understanding into the conditions under which delay terms can arise. By providing the conditions under which the delay exponential function is a probability distribution, we establish a critical value for the delay terms. An exact stochastic simulation method is introduced for the generalized model, enabling us to utilize the simulation in scenarios where intrinsic stochasticity is significant, such as when the population size is small. We illustrate the applications of the model and validate our simulation algorithm on examples drawn from epidemiology and pharmacokinetics., Comment: 20 pages, 4 figures

This paper introduces a novel incremental distributed back-end algorithm for Collaborative Simultaneous Localization and Mapping (C-SLAM). For real-world deployments, robotic teams require algorithms to compute a consistent state estimate accurately, within online runtime constraints, and with potentially limited communication. Existing centralized, decentralized, and distributed approaches to solving C-SLAM problems struggle to achieve all of these goals. To address this capability gap, we present Incremental Manifold Edge-based Separable ADMM (iMESA) a fully distributed C-SLAM back-end algorithm that can provide a multi-robot team with accurate state estimates in real-time with only sparse pair-wise communication between robots. Extensive evaluation on real and synthetic data demonstrates that iMESA is able to outperform comparable state-of-the-art C-SLAM back-ends., Comment: Accepted to Robotic Science and Systems (RSS) 2024 in Delft, Netherlands

Containerized vertical farming is a type of vertical farming practice using hydroponics in which plants are grown in vertical layers within a mobile shipping container. Space limitations within shipping containers make the automation of different farming operations challenging. In this paper, we explore the use of cobots (i.e., collaborative robots) to automate two key farming operations, namely, the transplantation of saplings and the harvesting of grown plants. Our method uses a single demonstration from a farmer to extract the motion constraints associated with the tasks, namely, transplanting and harvesting, and can then generalize to different instances of the same task. For transplantation, the motion constraint arises during insertion of the sapling within the growing tube, whereas for harvesting, it arises during extraction from the growing tube. We present experimental results to show that using RGBD camera images (obtained from an eye-in-hand configuration) and one demonstration for each task, it is feasible to perform transplantation of saplings and harvesting of leafy greens using a cobot, without task-specific programming.

In this paper we present a fully distributed, asynchronous, and general purpose optimization algorithm for Consensus Simultaneous Localization and Mapping (CSLAM). Multi-robot teams require that agents have timely and accurate solutions to their state as well as the states of the other robots in the team. To optimize this solution we develop a CSLAM back-end based on Consensus ADMM called MESA (Manifold, Edge-based, Separable ADMM). MESA is fully distributed to tolerate failures of individual robots, asynchronous to tolerate communication delays and outages, and general purpose to handle any CSLAM problem formulation. We demonstrate that MESA exhibits superior convergence rates and accuracy compare to existing state-of-the art CSLAM back-end optimizers., Comment: Accepted to ICRA 2024

This paper presents a method for robust optimization for online incremental Simultaneous Localization and Mapping (SLAM). Due to the NP-Hardness of data association in the presence of perceptual aliasing, tractable (approximate) approaches to data association will produce erroneous measurements. We require SLAM back-ends that can converge to accurate solutions in the presence of outlier measurements while meeting online efficiency constraints. Existing robust SLAM methods either remain sensitive to outliers, become increasingly sensitive to initialization, or fail to provide online efficiency. We present the robust incremental Smoothing and Mapping (riSAM) algorithm, a robust back-end optimizer for incremental SLAM based on Graduated Non-Convexity. We demonstrate on benchmarking datasets that our algorithm achieves online efficiency, outperforms existing online approaches, and matches or improves the performance of existing offline methods., Comment: Accepted to ICRA 2023

Reaction-diffusion equations are widely used as the governing evolution equations for modeling many physical, chemical, and biological processes. Here we derive reaction-diffusion equations to model transport with reactions on a one-dimensional domain that is evolving. The model equations, which have been derived from generalized continuous time random walks, can incorporate complexities such as subdiffusive transport and inhomogeneous domain stretching and shrinking. A method for constructing analytic expressions for short time moments of the position of the particles is developed and moments calculated from this approach are shown to compare favourably with results from random walk simulations and numerical integration of the reaction transport equation. The results show the important role played by the initial condition. In particular, it strongly affects the time dependence of the moments in the short time regime by introducing additional drift and diffusion terms. We also discuss how our reaction transport equation could be applied to study the spreading of a population on an evolving interface., Comment: 38 pages, 10 figures

Many physical phenomena occur on domains that grow in time. When the timescales of the phenomena and domain growth are comparable, models must include the dynamics of the domain. A widespread intrinsically slow transport process is subdiffusion. Many models of subdiffusion include a history dependence. This greatly confounds efforts to incorporate domain growth. Here we derive the fractional partial differential equations that govern subdiffusion on a growing domain, based on a Continuous Time Random Walk. This requires the introduction of a new, comoving, fractional derivative., Comment: 12 pages, 1 figure

Explicit numerical finite difference schemes for partial differential equations are well known to be easy to implement but they are particularly problematic for solving equations whose solutions admit shocks, blowups and discontinuities. Here we present an explicit numerical scheme for solving non-linear advection-diffusion equations admitting shock solutions that is both easy to implement and stable. The numerical scheme is obtained by considering the continuum limit of a discrete time and space stochastic process for non-linear advection-diffusion. The stochastic process is well posed and this guarantees the stability of the scheme. Several examples are provided to highlight the importance of the formulation of the stochastic process in obtaining a stable and accurate numerical scheme., Comment: 15 pages, 2 figures

There has recently been considerable interest in using a nonstandard piecewise approximation to formulate fractional order differential equations as difference equations that describe the same dynamical behaviour and are more amenable to a dynamical systems analysis. Unfortunately, due to mistakes in the fundamental papers, the difference equations formulated through this process do not capture the dynamics of the fractional order equations. We show that the correct application of this nonstandard piecewise approximation leads to a one parameter family of fractional order differential equations that converges to the original equation as the parameter tends to zero. A closed formed solution exists for each member of this family and leads to the formulation of a difference equation that is of increasing order as time steps are taken. Whilst this does not lead to a simplified dynamical analysis it does lead to a numerical method for solving the fractional order differential equation. The method is shown to be equivalent to a quadrature based method, despite the fact that it has not been derived from a quadrature. The method can be implemented with non-uniform time steps. An example is provided showing that the difference equation can correctly capture the dynamics of the underlying fractional differential equation.

Fractional-order SIR models have become increasingly popular in the literature in recent years, however unlike the standard SIR model, they often lack a derivation from an underlying stochastic process. Here we derive a fractional-order infectivity SIR model from a stochastic process that incorporates a time-since-infection dependence on the infectivity of individuals. The fractional derivative appears in the generalised master equations of a continuous time random walk through SIR compartments, with a power-law function in the infectivity. We show that this model can also be formulated as an infection-age structured Kermack-McKendrick integro-differential SIR model. Under the appropriate limit the fractional infectivity model reduces to the standard ordinary differential equation SIR model., Comment: 16 pages, no figures

Over the past several decades there has been a proliferation of epidemiological models with ordinary derivatives replaced by fractional derivatives in an an-hoc manner. These models may be mathematically interesting but their relevance is uncertain. Here we develop an SIR model for an epidemic, including vital dynamics, from an underlying stochastic process. We show how fractional differential operators arise naturally in these models whenever the recovery time from the disease is power law distributed. This can provide a model for a chronic disease process where individuals who are infected for a long time are unlikely to recover. The fractional order recovery model is shown to be consistent with the Kermack-McKendrick age-structured SIR model and it reduces to the Hethcote-Tudor integral equation SIR model. The derivation from a stochastic process is extended to discrete time, providing a stable numerical method for solving the model equations. We have carried out simulations of the fractional order recovery model showing convergence to equilibrium states. The number of infecteds in the endemic equilibrium state increases as the fractional order of the derivative tends to zero., Comment: 32 pages, 3 figures

BOUT++ is a 3D nonlinear finite-difference plasma simulation code, capable of solving quite general systems of PDEs, but targeted particularly on studies of the edge region of tokamak plasmas. BOUT++ is publicly available, and has been adopted by a growing number of researchers worldwide. Here we present improvements which have been made to the code since its original release, both in terms of structure and its capabilities. Some recent applications of these methods are reviewed, and areas of active development are discussed. We also present algorithms and tools which have been developed to enable creation of inputs from analytic expressions and experimental data, and for processing and visualisation of output results. This includes a new tool Hypnotoad for the creation of meshes from experimental equilibria. Algorithms have been implemented in BOUT++ to solve a range of linear algebraic problems encountered in the simulation of reduced MHD and gyro-fluid models: A preconditioning scheme is presented which enables the plasma potential to be calculated efficiently using iterative methods supplied by the PETSc library, without invoking the Boussinesq approximation. Scaling studies are also performed of a linear solver used as part of physics-based preconditioning to accelerate the convergence of implicit time-integration schemes., Comment: 16 pages, 8 figures, submitted to Journal of Plasma Physics

We describe the construction of stepped-pressure equilibria as extrema of a multi-region, relaxed magnetohydrodynamic (MHD) energy functional that combines elements of ideal MHD and Taylor relaxation, and which we call MRXMHD. The model is compatible with Hamiltonian chaos theory and allows the three-dimensional MHD equilibrium problem to be formulated in a well-posed manner suitable for computation. The energy-functional is discretized using a mixed finite-element, Fourier representation for the magnetic vector potential and the equilibrium geometry; and numerical solutions are constructed using the stepped-pressure equilibrium code, SPEC. Convergence studies with respect to radial and Fourier resolution are presented.

We describe a magnetohydrodynamic (MHD) constrained energy functional for equilibrium calculations that combines the topological constraints of ideal MHD with elements of Taylor relaxation. Extremizing states allow for partially chaotic magnetic fields and non-trivial pressure profiles supported by a discrete set of ideal interfaces with irrational rotational transforms. Numerical solutions are computed using the Stepped Pressure Equilibrium Code, SPEC, and benchmarks and convergence calculations are presented., Comment: Submitted to Plasma Physics and Controlled Fusion for publication with a cluster of papers associated with workshop: Stability and Nonlinear Dynamics of Plasmas, October 31, 2009 Atlanta, GA on occasion of 65th birthday of R.L. Dewar. V2 is revised for referees

The vanishing of the divergence of the total stress tensor (magnetic plus kinetic) in a neighborhood of an equilibrium plasma containing a toroidal surface of discontinuity gives boundary and jump conditions that strongly constrain allowable continuations of the magnetic field across the surface. The boundary conditions allow the magnetic fields on either side of the discontinuity surface to be described by surface magnetic potentials, reducing the continuation problem to that of solving a Hamilton--Jacobi equation. The characteristics of this equation obey Hamiltonian equations of motion, and a necessary condition for the existence of a continued field across a general toroidal surface is that there exist invariant tori in the phase space of this Hamiltonian system. It is argued from the Birkhoff theorem that existence of such an invariant torus is also, in general, sufficient for continuation to be possible. An important corollary is that the rotational transform of the continued field on a surface of discontinuity must, generically, be irrational., Comment: Prepared for submission to Phys. Letts. A

The flexible Heliac coil set of helical axis stellarator H-1 (R=1m, ~0.15-0.2 m) permits access to a wide range of magnetic configurations. Surprisingly, in the absence of any obvious population of energetic particles, Alfven modes normally associated with energetic populations in larger fusion experiments are observed. Using H-1's unique combination of flexibility and advanced diagnostics, RF-generated plasma in H-1 is shown to have a very complex dependence on configuration of both the electron density and fluctuations in the MHD Alfven range. Magnetic fluctuations range from highly coherent, often multi-frequency, to approaching broad-band (df/f ~ 0.02-0.5), in the range 1-200 kHz. Application of datamining techniques to a wide range of configurations classifies these fluctuations and extracts poloidal and toroidal mode numbers, revealing that a significant class of fluctuations exhibit scaling which is i) Alfvenic with electron density (within a constant factor) and ii) shear Alfvenic in rotational transform. This is confirmed by scans within a single pulse, which can follow mode conversions. An array of optical and interferometric diagnostics is combined with the magnetic probe arrays to provide initial information on the internal structure of the MHD modes, and associated 3D effects. The configurational dependence is closely related to the presence of low order rational surfaces; density falls to very low values near, but not precisely at these rational values. Results from a uniquely accurate magnetic field mapping system, combined with a comprehensive model of the vacuum magnetic field in H-1 show that magnetic islands should not dominate the confinement of the configuration, and indicate that the dependence of density on configuration may be attributable to plasma generation effects., Comment: Expanded form of paper Ex/P9-11 presented at the 22nd IAEA Fusion Energy Conference, Geneva, Switzerland, 2008, and submitted to Nuclear Fusion for Publication

The concept of plasma relaxation as a constrained energy minimization is reviewed. Recent work by the authors on generalizing this approach to partially relaxed three-dimensional plasma systems in a way consistent with chaos theory is discussed, with a view to clarifying the thermodynamic aspects of the variational approach used. Other entropy-related approaches to finding long-time steady states of turbulent or chaotic plasma systems are also briefly reviewed., Comment: Contribution to the Proceedings of the AMSI/MASCOS Concepts of Entropy and their Applications Workshop, Melbourne, Australia, November-December 2007, to be published in Entropy e-journal http://www.mdpi.org/entropy/