Search

Autonomous surgical robots are a promising solution to the increasing demand for surgery amid a shortage of surgeons. Recent work has proposed learning-based approaches for the autonomous manipulation of soft tissue. However, due to variability in tissue geometries and stiffnesses, these methods do not always perform optimally, especially in out-of-distribution settings. We propose, develop, and test the first application of uncertainty quantification to learned surgical soft-tissue manipulation policies as an early identification system for task failures. We analyze two different methods of uncertainty quantification, deep ensembles and Monte Carlo dropout, and find that deep ensembles provide a stronger signal of future task success or failure. We validate our approach using the physical daVinci Research Kit (dVRK) surgical robot to perform physical soft-tissue manipulation. We show that we are able to successfully detect task failure and request human intervention when necessary while still enabling autonomous manipulation when possible. Our learned tissue manipulation policy with uncertainty-based early failure detection achieves a zero-shot sim2real performance improvement of 47.5% over the prior state of the art in learned soft-tissue manipulation. We also show that our method generalizes well to new types of tissue as well as to a bimanual soft tissue manipulation task., Comment: 8 pages, 6 figures

Satellite to ground quantum communication typically operates at night to reduce background signals, however it remains susceptible to noise from light pollution of the night sky. In this study we compare several methodologies for determining whether a Quantum Ground Station (QGS) site is viable for exchanging quantum signals with the upcoming Quantum Encryption and Science Satellite (QEYSSat) mission. We conducted ground site characterization studies at three locations in Canada: Waterloo, Ontario, Calgary, Alberta, and Priddis, Alberta. Using different methods we estimate the background counts expected to leak into the satellite-ground quantum channel, and determined whether the noise levels could prevent a quantum key transfer. We also investigate how satellite data recorded from the Visible Infrared Imaging Radiometer Suite (VIIRS) can help estimate conditions of a particular site, and find reasonable agreement with the locally recorded data. Our results indicate that the Waterloo, Calgary, and Priddis QGS sites should allow both quantum uplinks and downlinks with QEYSSat, despite their proximity to urban centres. Furthermore, our approach allows the use of satellite borne instrument data (VIIRS) to remotely and efficiently determine the potential of a ground site., Comment: Submitted to EPJ: Prospects for Space Quantum Research, 20 pages, 12 figures

We present the first general stability results for nonlinear offset-free model predictive control (MPC). Despite over twenty years of active research, the offset-free MPC literature has not shaken the assumption of closed-loop stability for establishing offset-free performance. In this paper, we present a nonlinear offset-free MPC design that is robustly stable with respect to the tracking errors, and thus achieves offset-free performance, despite plant-model mismatch and persistent disturbances. Key features and assumptions of this design include quadratic costs, differentiability of the plant and model functions, constraint backoffs at steady state, and a robustly stable state and disturbance estimator. We first establish nominal stability and offset-free performance. Then, robustness to state and disturbance estimate errors and setpoint and disturbance changes is demonstrated. Finally, the results are extended to sufficiently small plant-model mismatch. The results are illustrated by numerical examples., Comment: 55 pages, 4 figures

We consider the asymptotic stability of MPC under plant-model mismatch, considering primarily models where the origin remains a steady state despite mismatch. Our results differ from prior results on the inherent robustness of MPC, which guarantee only convergence to a neighborhood of the origin, the size of which scales with the magnitude of the mismatch. For MPC with quadratic costs, continuous differentiability of the system dynamics is sufficient to demonstrate exponential stability of the closed-loop system despite mismatch. For MPC with general costs, a joint comparison function bound and scaling condition guarantee asymptotic stability despite mismatch. The results are illustrated in both algebraic and engineering examples. The tools developed to establish these results can address the stability of offset-free MPC, an open and interesting question in the MPC research literature., Comment: 42 pages, 9 figures

The significance of statistical physics concepts such as entropy extends far beyond classical thermodynamics. We interpret the similarity between partitions in statistical mechanics and partitions in Bayesian inference as an articulation of a result by Jaynes (1957), who clarified that thermodynamics is in essence a theory of information. In this, every sampling process has a mechanical analogue. Consequently, the divide between ensembles of samplers in parameter space and sampling from a mechanical system in thermodynamic equilibrium would be artificial. Based on this realisation, we construct a continuous modelling of a Bayes update akin to a transition between thermodynamic ensembles. This leads to an information theoretic interpretation of Jazinsky's equality, relating the expenditure of work to the influence of data via the likelihood. We propose one way to transfer the vocabulary and the formalism of thermodynamics (energy, work, heat) and statistical mechanics (partition functions) to statistical inference, starting from Bayes' law. Different kinds of inference processes are discussed and relative entropies are shown to follow from suitably constructed partitions as an analytical formulation of sampling processes. Lastly, we propose an effective dimension as a measure of system complexity. A numerical example from cosmology is put forward to illustrate these results.

We explore scalar radiation from neutron star-black hole binaries in scalar-tensor theories with kinetic screening ($K$-essence). Using 3+1 numerical relativity simulations in the decoupling limit, we investigate scalar dipole and quadrupole radiation for different values of the strong coupling constant $\Lambda$. Our results show that kinetic screening effectively suppresses the scalar dipole radiation as $\Lambda$ decreases. This is validated by comparing to analytic predictions for the screening of dipole scalar emission, with which our numerical results show good agreement. However, our numerical simulations show that the suppression of scalar quadrupole radiation is less efficient, even when the screening radius exceeds the wavelength of the emitted radiation. In fact, the dependence of the scalar quadrupole amplitude on $\Lambda$ flattens out for the smallest $\Lambda$ that we can simulate, and the quadrupole amplitude is suppressed only by a factor $\lesssim 3$ relative to the Fierz-Jordan-Brans-Dicke case. Overall, our study shows that scalar quadrupole radiation from mixed binaries may be used to place constraints on $K$-essence theories with next-generation gravitational-wave detectors., Comment: 16 pages, 13 figures

Theories described by non-Hermitian Hamiltonians are known to possess strictly positive energy eigenvalues and exhibit unitary time evolution if the Hamiltonian is symmetric under discrete parity and time (PT) transformation. In this work, we demonstrate how quantum quadratic gravity, a theory that generally violates unitarity when viewed as a Hermitian quantum field theory, can be complex-deformed into such a PT-symmetric theory with an action that consists of a ghost-less Hermitian free part and non-Hermitian interactions. Paying special attention to the gauge symmetry present in the theory, we quantize in the covariant operator formalism after suggesting how the framework might be extended to the pseudo-Hermitian picture. We find compelling evidence that the resulting quantum theory possesses a unitary inner product and a sensible interpretation of quantum probability, thus avoiding the ghost problem present in the Hermitian formulation of quadratic gravity., Comment: 32+4 pages, v2: typos fixed and minor changes to discussion

It is well-known that the results by Bekenstein, Gibbons and Hawking on the thermodynamics of black holes can be reproduced quite simply in the Euclidean path integral approach to Quantum Gravity. The corresponding partition function is obtained semiclassically, ultimately requiring only the on-shell Einstein-Hilbert action with opportune asymptotic subtractions. We elaborate on the fact that the same expressions for the thermodynamical quantities can be obtained within teleparallel equivalent theories, based on either torsion or nonmetricity, by employing quasilocal relations. Notably, the bulk integrals of these theories do not vanish on-shell but rather result in boundary terms themselves. Asymptotic subtractions of the latter are able to cancel out the divergences, ultimately leading to Bekenstein-Gibbons-Hawking's results. As a non-trivial cross-check, we compute the bulk integrals directly without reference to the boundary terms. While the result agrees with the previous method for the torsion-based teleparallel theory, it differs for the nonmetricity theory. Specifically, upon regularizing the bulk integral using a fiducial reference frame, we find that the semiclassical partition function vanishes. To address this problem, we propose a simple prescription for Schwarzschild black holes, which involves keeping the nonmetric connection arbitrary and imposing thermal equilibrium. Generalizations of the results to more general modified gravity theories with antisymmetric degrees of freedom are also discussed., Comment: 21 pages, improved version to appear in CQG

Using novel metallic magnetic calorimeter detectors at the CRYRING@ESR, we recorded X-ray spectra of stored and electron cooled helium-like uranium (U$^{90+}$) with an unmatched spectral resolution of close to 90 eV. This allowed for an accurate determination of the energies of all four components of the K$\alpha$ transitions in U$^{90+}$. We find good agreement with state-of-the-art bound-state QED calculations for the strong-field regime. Our results do not support any systematic deviation between experiment and theory in helium-like systems, the presence of which was subject of intense debates in recent years.

Autonomous robotic inspection, where a robot moves through its environment and inspects points of interest, has applications in industrial settings, structural health monitoring, and medicine. Planning the paths for a robot to safely and efficiently perform such an inspection is an extremely difficult algorithmic challenge. In this work we consider an abstraction of the inspection planning problem which we term Graph Inspection. We give two exact algorithms for this problem, using dynamic programming and integer linear programming. We analyze the performance of these methods, and present multiple approaches to achieve scalability. We demonstrate significant improvement both in path weight and inspection coverage over a state-of-the-art approach on two robotics tasks in simulation, a bridge inspection task by a UAV and a surgical inspection task using a medical robot.

The purpose of this note is to summarize the arguments required to derive the results appearing in robust minmax control of linear dynamical systems using a quadratic stage cost. The main results required in robust minmax control are Corollary 19 and Proposition 20. Moreover, the solution to the trust-region problem given in Proposition 15 and Lemma 16 may be of more general interest.

General Relativity predicts the existence of quadratic quasi-normal modes at second order in perturbation theory. Building on our recent work, we compute the amplitudes and polarizations of these modes for non-rotating black holes, showing that they are completely determined by the amplitudes and polarizations of linear modes. We obtain the ratio of quadratic to linear amplitudes, which still depends on the initial conditions of the merger through the polarization of linear modes. However, we demonstrate that this dependence is captured by four fundamental numbers, independent of initial conditions, representing four different combinations of linear modes parities. Additionally, we prove two selection rules regarding the vanishing of classes of quadratic modes. Our results are available online as a package which provides the ratio of amplitudes across a broad spectrum of angular momenta.

This paper addresses the identification of models for offset-free model predictive control (MPC), where LTI models are augmented with (fictitious) uncontrollable integrating modes, called integrating disturbances. The states and disturbances are typically estimated with a Kalman filter. The disturbance estimates effectively provide integral control, so the quality of the disturbance model (and resulting filter) directly influences the control performance. We implement eigenvalue constraints to protect against undesirable filter behavior (unstable or marginally stable modes, high-frequency oscillations). Specifically, we consider the class of linear matrix inequality (LMI) regions for eigenvalue constraints. These LMI regions are open sets by default, so we introduce a barrier function method to create tightened, but closed, eigenvalue constraints. To solve the resulting nonlinear semidefinite program, we approximate it as a nonlinear program using a Cholesky factorization method that exploits known sparsity structures of semidefinite optimization variables and matrix inequalities. The algorithm is applied to real-world data taken from two physical systems: first, a low-cost benchmark temperature microcontroller suitable for classroom laboratories, and second, an industrial-scale chemical reactor at Eastman Chemical's plant in Kingsport, TN., Comment: 21 pages, 14 figures

Besides classical feed-forward neural networks, also neural ordinary differential equations (neural ODEs) gained particular interest in recent years. Neural ODEs can be interpreted as an infinite depth limit of feed-forward or residual neural networks. We study the input-output dynamics of finite and infinite depth neural networks with scalar output. In the finite depth case, the input is a state associated to a finite number of nodes, which maps under multiple non-linear transformations to the state of one output node. In analogy, a neural ODE maps a linear transformation of the input to a linear transformation of its time-$T$ map. We show that depending on the specific structure of the network, the input-output map has different properties regarding the existence and regularity of critical points. These properties can be characterized via Morse functions, which are scalar functions, where every critical point is non-degenerate. We prove that critical points cannot exist, if the dimension of the hidden layer is monotonically decreasing or the dimension of the phase space is smaller or equal to the input dimension. In the case that critical points exist, we classify their regularity depending on the specific architecture of the network. We show that each critical point is non-degenerate, if for finite depth neural networks the underlying graph has no bottleneck, and if for neural ODEs, the linear transformations used have full rank. For each type of architecture, the proven properties are comparable in the finite and in the infinite depth case. The established theorems allow us to formulate results on universal embedding, i.e.\ on the exact representation of maps by neural networks and neural ODEs. Our dynamical systems viewpoint on the geometric structure of the input-output map provides a fundamental understanding, why certain architectures perform better than others.

Quadratic quasi-normal modes, generated at second order in black hole perturbation theory, are a promising target for testing gravity in the nonlinear regime with next-generation gravitational wave detectors. While their frequencies have long been known, their amplitudes remain poorly studied. We introduce regular variables and compute amplitudes for Schwarzschild black holes with the Leaver algorithm. We find a nonlinear ratio $\mathcal{R}\simeq0.154e^{-0.068i}$ for the most excited $\ell=4$ mode, matching results from Numerical Relativity. We also predict new low-frequency $\ell=2$ quadratic modes., Comment: 5 pages, 2 figures. v2: matches published version

Millisecond pulsars are extremely precise celestial clocks: as they rotate, the beamed radio waves emitted along the axis of their magnetic field can be detected with radio telescopes, which allows for tracking subtle changes in the pulsars' rotation periods. A possible effect on the period of a pulsar is given by a potential coupling to dark matter, in cases where it is modeled with an "ultralight" scalar field. In this paper, we consider a universal conformal coupling of the dark matter scalar to gravity, which in turn mediates an effective coupling between pulsars and dark matter. If the dark matter scalar field is changing in time, as expected in the Milky Way, this effective coupling produces a periodic modulation of the pulsar rotational frequency. By studying the time series of observed radio pulses collected by the European Pulsar Timing Array experiment, we present constraints on the coupling of dark matter, improving on existing bounds. These bounds can also be regarded as constraints on the parameters of scalar-tensor theories of the Fierz-Jordan-Brans-Dicke and Damour-Esposito-Far\`{e}se types in the presence of a (light) mass potential term., Comment: 17 pages, 5 figures, Updated plots, matching published version

Automating robotic surgery via learning from demonstration (LfD) techniques is extremely challenging. This is because surgical tasks often involve sequential decision-making processes with complex interactions of physical objects and have low tolerance for mistakes. Prior works assume that all demonstrations are fully observable and optimal, which might not be practical in the real world. This paper introduces a sample-efficient method that learns a robust reward function from a limited amount of ranked suboptimal demonstrations consisting of partial-view point cloud observations. The method then learns a policy by optimizing the learned reward function using reinforcement learning (RL). We show that using a learned reward function to obtain a policy is more robust than pure imitation learning. We apply our approach on a physical surgical electrocautery task and demonstrate that our method can perform well even when the provided demonstrations are suboptimal and the observations are high-dimensional point clouds. Code and videos available here: https://sites.google.com/view/lfdinelectrocautery, Comment: In proceedings of the International Symposium on Medical Robotics (ISMR) 2024. Equal contribution from two first authors

By considering an arbitrary bare action describing BSM physics, we use the Barvinsky-Vilkovisky resummation to obtain the most general non-local electroweak effective action at second order in the field strength. We also include the contribution of the functional measure to the effective action, which is found to modify the Higgs potential by shifting its vacuum value. The resulting effective action provides one-loop corrections to the $W$ and $Z$ boson masses, ultimately leading to the most general expression for the $\rho$ parameter at one-loop. The functional measure plays a pivotal role as it allows the parameterization of $\rho$ in inverse powers of the scale of new physics, while containing non-local form factors. The comparison of $\rho$ with the latest data leads to several constraints on the UV particle spectra of BSM models., Comment: Matches the version accepted in JHEP

Tendon-driven continuum robot kinematic models are frequently computationally expensive, inaccurate due to unmodeled effects, or both. In particular, unmodeled effects produce uncertainties that arise during the robot's operation that lead to variability in the resulting geometry. We propose a novel solution to these issues through the development of a Gaussian mixture kinematic model. We train a mixture density network to output a Gaussian mixture model representation of the robot geometry given the current tendon displacements. This model computes a probability distribution that is more representative of the true distribution of geometries at a given configuration than a model that outputs a single geometry, while also reducing the computation time. We demonstrate one use of this model through a trajectory optimization method that explicitly reasons about the workspace uncertainty to minimize the probability of collision.

Tendon-driven continuum robots have been gaining popularity in medical applications due to their ability to curve around complex anatomical structures, potentially reducing the invasiveness of surgery. However, accurate modeling is required to plan and control the movements of these flexible robots. Physics-based models have limitations due to unmodeled effects, leading to mismatches between model prediction and actual robot shape. Recently proposed learning-based methods have been shown to overcome some of these limitations but do not account for hysteresis, a significant source of error for these robots. To overcome these challenges, we propose a novel deep decoder neural network that predicts the complete shape of tendon-driven robots using point clouds as the shape representation, conditioned on prior configurations to account for hysteresis. We evaluate our method on a physical tendon-driven robot and show that our network model accurately predicts the robot's shape, significantly outperforming a state-of-the-art physics-based model and a learning-based model that does not account for hysteresis., Comment: 8 pages, 9 figures, Submitted to IEEE Robotics and Automation Letters

NGC 4945 contains a well-known heavily obscured active galactic nucleus (AGN) at its core, with prior reports of strong nuclear and off-nuclear neutral Fe K$\alpha$ emission due to the AGN activity. We report the discovery of very extended Fe K$\alpha$ emission with the XMM-Newton EPIC pn in a $\sim5$ kpc by $\sim10$ kpc region that is misaligned with the plane of the inclined optical galaxy disk by $\sim60$ degrees in projection. After a careful consideration of the crowded center of the galaxy and numerous unresolved hard X-ray sources present, we estimate that $\sim15$% of the Fe K$\alpha$ is extended on kpc-sized scales. The overall size and misalignment of the region follows an unusual pattern of radio polarization that is not typical of starbursts or normal disk galaxies but has been interpreted as possibly due to AGN activity. We suggest that the extended Fe K$\alpha$ emission arose from a period of AGN eruption several million years ago - a relic of a past AGN ejection episode., Comment: 24 pages including figures and tables, 12 figures, 3 tables. Submitted to ApJ, under review

The effective action for quantum gravity coupled to matter contains corrections arising from the functional measure. We analyse the effect of such corrections for anisotropic self-gravitating compact objects described by means of the gravitational decoupling method applied to isotropic solutions of the Einstein field equations. In particular, we consider the Tolman IV solution of general relativity and show that quantum gravity effects can modify the effective energy density as well as the effective tangential and radial pressures. For a suitable choice of the mimicking constant, upper bounds on the quantum corrections can be driven by the surface redshift of the anisotropic compact stellar system obtained with the gravitational decoupling., Comment: 30 pages, 10 figures

In the aftermath of a binary black hole merger event, the gravitational wave signal emitted by the remnant black hole is modeled as a superposition of damped sinusoids known as quasinormal modes. While the dominant quasinormal modes originating from linear black hole perturbation theory have been studied extensively in this post-merger "ringdown" phase, more accurate models of ringdown radiation include the nonlinear modes arising from higher-order perturbations of the remnant black hole spacetime. We explore the detectability of quadratic quasinormal modes with both ground- and space-based next-generation detectors. We quantify how predictions of the quadratic mode detectability depend on the quasinormal mode starting times. We then calculate the signal-to-noise ratio of quadratic modes for several detectors and binary black hole populations, focusing on the ($220\times220$) mode - i.e., on the quadratic term sourced by the square of the linear $(220)$ mode. For the events with the loudest quadratic mode signal-to-noise ratios, we additionally compute statistical errors on the mode parameters in order to further ascertain the distinguishability of the quadratic mode from the linear quasinormal modes. The astrophysical models used in this paper suggest that while the quadratic mode may be detectable in at most a few events with ground-based detectors, the prospects for detection with the Laser Interferometer Space Antenna (LISA) are more optimistic., Comment: 21 pages, 16 figures, 2 tables; revised to match published version; new plots in App. D

Scalar-tensor theories have a long history as possible phenomenological alternatives to General Relativity, but are known to potentially produce deviations from the (strong) equivalence principle in systems involving self-gravitating objects, as a result of the presence of an additional gravitational scalar field besides the tensor modes of General Relativity. We describe here a novel mechanism whereby the equivalence principle is violated for an isolated rotating neutron star, if the gravitational scalar field is changing in time far from the system. We show that the neutron star rotational period changes due to an effective coupling ("angular momentum sensitivity") to the gravitational scalar, and compute that coupling for viable equations of state for nuclear matter. We comment on the relevance of our findings for testing scalar-tensor theories and models of ultralight dark matter with pulsar timing observations, a topic that we tackle in a companion paper.

We prove non-asymptotic error bounds for particle gradient descent (PGD)~(Kuntz et al., 2023), a recently introduced algorithm for maximum likelihood estimation of large latent variable models obtained by discretizing a gradient flow of the free energy. We begin by showing that, for models satisfying a condition generalizing both the log-Sobolev and the Polyak--{\L}ojasiewicz inequalities (LSI and P{\L}I, respectively), the flow converges exponentially fast to the set of minimizers of the free energy. We achieve this by extending a result well-known in the optimal transport literature (that the LSI implies the Talagrand inequality) and its counterpart in the optimization literature (that the P{\L}I implies the so-called quadratic growth condition), and applying it to our new setting. We also generalize the Bakry--\'Emery Theorem and show that the LSI/P{\L}I generalization holds for models with strongly concave log-likelihoods. For such models, we further control PGD's discretization error, obtaining non-asymptotic error bounds. While we are motivated by the study of PGD, we believe that the inequalities and results we extend may be of independent interest.

We discuss an extra-dimensional braneworld with a 5th dimension compactified on a circle. As a characteristic feature, the warp factor is hyperbolic and separates the hidden and visible branes by a bulk horizon without a singularity. The two most widely separated scales of 4D physics - the 4D Planck mass and 4D cosmological constant - are determined by two physical scales in the extra dimension, namely: $(i)$ the proper size of the extra dimension, $R$, and, $(ii)$ the distance between the visible brane and the horizon, $R_0$. A realistic scale hierarchy between 4D Planck mass and 4D cosmological constant is obtained for $R/R_0\sim2.34$. The usual fine tuning is not reduced but promoted to a fine tuning of two separate brane energy densities that must approach the fundamental scale of the model with very high precision. Our scenario is based on an exact solution to the 5D Einstein equations with a strictly empty bulk and Friedmann-Lema\^itre-Robertson-Walker metric on the 4D branes. This requires positive 4D brane energy densities and describes an adiabatic runaway solution in agreement with the de Sitter swampland conjecture. The Kaluza-Klein (KK) graviton states are solutions of a modified P\"oschl-Teller potential which permits a discrete graviton spectrum of $\textit{exactly two}$ modes. In addition to the usual massless graviton, our scenario predicts an extra massive spin-2 graviton with a mass gap of $m_1=\sqrt{2}H_0\approx2\times10^{-33}\,\mathrm{eV}$ which might be detectable in the foreseeable future. A KK tower of gravitons, or a possible continuum of massive graviton states, is prohibited by unitarity with respect to the horizon. We discuss hurdles in turning this model into a realistic cosmology at all times, which points us towards 4D brane tensions that that must be raising towards the fundamental scale of the model, while the observable 4D expansion rate is decreasing., Comment: 33 pages, 2 figures

Living objects are able to consume chemical energy and process information independently from others. However, living objects can coordinate to form ordered groups such as schools of fish. This work considers these complex groups as living materials and presents imaging-based experiments of laboratory schools of fish to understand how this non-equilibrium activity affects the mechanical properties of a group. We use spatial confinement to control the motion and structure of fish within quasi-2D shoals of fish. Using image analysis techniques, we make quantitative observations of the structures, their spatial heterogeneity, and their temporal fluctuations. Furthermore, we utilize Monte Carlo simulations to replicate the experimentally observed area distribution patterns which provide insight into the effective interactions between fish and confirm the presence of a confinement-based behavioral preference transition. In addition, unlike in short-range interacting systems, here structural heterogeneity and dynamic activities are positively correlated as a result of complex interplay between spatial arrangement and behavioral dynamics in fish collectives., Comment: 18 pages, 7 Figures

Maximum likelihood estimation (MLE) of latent variable models is often recast as the minimization of a free energy functional over an extended space of parameters and probability distributions. This perspective was recently combined with insights from optimal transport to obtain novel particle-based algorithms for fitting latent variable models to data. Drawing inspiration from prior works which interpret `momentum-enriched' optimization algorithms as discretizations of ordinary differential equations, we propose an analogous dynamical-systems-inspired approach to minimizing the free energy functional. The result is a dynamical system that blends elements of Nesterov's Accelerated Gradient method, the underdamped Langevin diffusion, and particle methods. Under suitable assumptions, we prove that the continuous-time system minimizes the functional. By discretizing the system, we obtain a practical algorithm for MLE in latent variable models. The algorithm outperforms existing particle methods in numerical experiments and compares favourably with other MLE algorithms., Comment: ICML 2024 camera ready

Monte-Carlo techniques are standard numerical tools for exploring non-Gaussian and multivariate likelihoods. Many variants of the original Metropolis-Hastings algorithm have been proposed to increase the sampling efficiency. Motivated by Ensemble Monte Carlo we allow the number of Markov chains to vary by exchanging particles with a reservoir, controlled by a parameter analogous to a chemical potential $\mu$, which effectively establishes a random process that samples microstates from a macrocanonical instead of a canonical ensemble. In this paper, we develop the theory of macrocanonical sampling for statistical inference on the basis of Bayesian macrocanonical partition functions, thereby bringing to light the relations between information-theoretical quantities and thermodynamic properties. Furthermore, we propose an algorithm for macrocanonical sampling, $\texttt{Avalanche Sampling}$, and apply it to various toy problems as well as the likelihood on the cosmological parameters $\Omega_m$ and $w$ on the basis of data from the supernova distance redshift relation., Comment: 13 pages, 9 figures