-
Making Robust Generalizers Less Rigid with Soft Ascent-Descent
Authors:
Matthew J. Holland,
Toma Hamada
Abstract:
While the traditional formulation of machine learning tasks is in terms of performance on average, in practice we are often interested in how well a trained model performs on rare or difficult data points at test time. To achieve more robust and balanced generalization, methods applying sharpness-aware minimization to a subset of worst-case examples have proven successful for image classification…
▽ More
While the traditional formulation of machine learning tasks is in terms of performance on average, in practice we are often interested in how well a trained model performs on rare or difficult data points at test time. To achieve more robust and balanced generalization, methods applying sharpness-aware minimization to a subset of worst-case examples have proven successful for image classification tasks, but only using deep neural networks in a scenario where the most difficult points are also the least common. In this work, we show how such a strategy can dramatically break down under more diverse models, and as a more robust alternative, instead of typical sharpness we propose and evaluate a training criterion which penalizes poor loss concentration, which can be easily combined with loss transformations such as CVaR or DRO that control tail emphasis.
△ Less
Submitted 7 August, 2024;
originally announced August 2024.
-
Entangled Matter-waves for Quantum Enhanced Sensing
Authors:
John Drew Wilson,
Jarrod T. Reilly,
Haoqing Zhang,
Chengyi Luo,
Anjun Chu,
James K. Thompson,
Ana Maria Rey,
Murray J. Holland
Abstract:
The ability to create and harness entanglement is crucial to the fields of quantum sensing and simulation, and ultracold atom-cavity systems offer pristine platforms for this undertaking. Here, we present a method for creating and controlling entanglement between solely the motional states of atoms in a cavity without the need for electronic interactions. We show this interaction arises from a gen…
▽ More
The ability to create and harness entanglement is crucial to the fields of quantum sensing and simulation, and ultracold atom-cavity systems offer pristine platforms for this undertaking. Here, we present a method for creating and controlling entanglement between solely the motional states of atoms in a cavity without the need for electronic interactions. We show this interaction arises from a general atom-cavity model, and discuss the role of the cavity frequency shift in response to atomic motion. This cavity response leads to many different squeezing interactions between the atomic momentum states. Furthermore, we show that when the atoms form a density grating, the collective motion leads to one-axis twisting, a many-body energy gap, and metrologically useful entanglement even in the presence of noise. Noteably, an experiment has recently demonstrated this regime leads to an effective momentum-exchange interaction between atoms in a common cavity mode. This system offers a highly tunable, many-body quantum sensor and simulator.
△ Less
Submitted 12 August, 2024; v1 submitted 19 June, 2024;
originally announced June 2024.
-
Robust Quantum Sensing with Multiparameter Decorrelation
Authors:
Shah Saad Alam,
Victor E. Colussi,
John Drew Wilson,
Jarrod T. Reilly,
Michael A. Perlin,
Murray J. Holland
Abstract:
The performance of a quantum sensor is fundamentally limited by noise. This noise is particularly damaging when it becomes correlated with the readout of a target signal, caused by fluctuations of the sensor's operating parameters. These uncertainties limit sensitivity in a way that can be understood with multiparameter estimation theory. We develop a new approach, adaptable to any quantum platfor…
▽ More
The performance of a quantum sensor is fundamentally limited by noise. This noise is particularly damaging when it becomes correlated with the readout of a target signal, caused by fluctuations of the sensor's operating parameters. These uncertainties limit sensitivity in a way that can be understood with multiparameter estimation theory. We develop a new approach, adaptable to any quantum platform, for designing robust sensing protocols that leverages multiparameter estimation theory and machine learning to decorrelate a target signal from fluctuating off-target (``nuisance'') parameters. Central to our approach is the identification of information-theoretic goals that guide a machine learning agent through an otherwise intractably large space of potential sensing protocols. As an illustrative example, we apply our approach to a reconfigurable optical lattice to design an accelerometer whose sensitivity is decorrelated from lattice depth noise. We demonstrate the effect of decorrelation on outcomes and Bayesian inferencing through statistical analysis in parameter space, and discuss implications for future applications in quantum metrology and computing.
△ Less
Submitted 13 May, 2024;
originally announced May 2024.
-
Criterion Collapse and Loss Distribution Control
Authors:
Matthew J. Holland
Abstract:
In this work, we consider the notion of "criterion collapse," in which optimization of one metric implies optimality in another, with a particular focus on conditions for collapse into error probability minimizers under a wide variety of learning criteria, ranging from DRO and OCE risks (CVaR, tilted ERM) to non-monotonic criteria underlying recent ascent-descent algorithms explored in the literat…
▽ More
In this work, we consider the notion of "criterion collapse," in which optimization of one metric implies optimality in another, with a particular focus on conditions for collapse into error probability minimizers under a wide variety of learning criteria, ranging from DRO and OCE risks (CVaR, tilted ERM) to non-monotonic criteria underlying recent ascent-descent algorithms explored in the literature (Flooding, SoftAD). We show how collapse in the context of losses with a Bernoulli distribution goes far beyond existing results for CVaR and DRO, then expand our scope to include surrogate losses, showing conditions where monotonic criteria such as tilted ERM cannot avoid collapse, whereas non-monotonic alternatives can.
△ Less
Submitted 21 May, 2024; v1 submitted 15 February, 2024;
originally announced February 2024.
-
Reductive Quantum Phase Estimation
Authors:
Nicholas J. C. Papadopoulos,
Jarrod T. Reilly,
John Drew Wilson,
Murray J. Holland
Abstract:
Estimating a quantum phase is a necessary task in a wide range of fields of quantum science. To accomplish this task, two well-known methods have been developed in distinct contexts, namely, Ramsey interferometry (RI) in atomic and molecular physics and quantum phase estimation (QPE) in quantum computing. We demonstrate that these canonical examples are instances of a larger class of phase estimat…
▽ More
Estimating a quantum phase is a necessary task in a wide range of fields of quantum science. To accomplish this task, two well-known methods have been developed in distinct contexts, namely, Ramsey interferometry (RI) in atomic and molecular physics and quantum phase estimation (QPE) in quantum computing. We demonstrate that these canonical examples are instances of a larger class of phase estimation protocols, which we call reductive quantum phase estimation (RQPE) circuits. Here we present an explicit algorithm that allows one to create an RQPE circuit. This circuit distinguishes an arbitrary set of phases with a fewer number of qubits and unitary applications, thereby solving a general class of quantum hypothesis testing to which RI and QPE belong. We further demonstrate a trade-off between measurement precision and phase distinguishability, which allows one to tune the circuit to be optimal for a specific application.
△ Less
Submitted 11 July, 2024; v1 submitted 6 February, 2024;
originally announced February 2024.
-
Implicit regularization via soft ascent-descent
Authors:
Matthew J. Holland,
Kosuke Nakatani
Abstract:
As models grow larger and more complex, achieving better off-sample generalization with minimal trial-and-error is critical to the reliability and economy of machine learning workflows. As a proxy for the well-studied heuristic of seeking "flat" local minima, gradient regularization is a natural avenue, and first-order approximations such as Flooding and sharpness-aware minimization (SAM) have rec…
▽ More
As models grow larger and more complex, achieving better off-sample generalization with minimal trial-and-error is critical to the reliability and economy of machine learning workflows. As a proxy for the well-studied heuristic of seeking "flat" local minima, gradient regularization is a natural avenue, and first-order approximations such as Flooding and sharpness-aware minimization (SAM) have received significant attention, but their performance depends critically on hyperparameters (flood threshold and neighborhood radius, respectively) that are non-trivial to specify in advance. In order to develop a procedure which is more resilient to misspecified hyperparameters, with the hard-threshold "ascent-descent" switching device used in Flooding as motivation, we propose a softened, pointwise mechanism called SoftAD that downweights points on the borderline, limits the effects of outliers, and retains the ascent-descent effect. We contrast formal stationarity guarantees with those for Flooding, and empirically demonstrate how SoftAD can realize classification accuracy competitive with SAM and Flooding while maintaining a much smaller loss generalization gap and model norm. Our empirical tests range from simple binary classification on the plane to image classification using neural networks with millions of parameters; the key trends are observed across all datasets and models studied, and suggest a potential new approach to implicit regularization.
△ Less
Submitted 15 October, 2023;
originally announced October 2023.
-
Speeding Up Squeezing with a Periodically Driven Dicke Model
Authors:
Jarrod T. Reilly,
Simon B. Jäger,
John Drew Wilson,
John Cooper,
Sebastian Eggert,
Murray J. Holland
Abstract:
We present a simple and effective method to create highly entangled spin states on a faster timescale than that of the commonly employed one-axis twisting (OAT) model. We demonstrate that by periodically driving the Dicke Hamiltonian at a resonance frequency, the system effectively becomes a two-axis countertwisting Hamiltonian which is known to quickly create Heisenberg limit scaled entangled sta…
▽ More
We present a simple and effective method to create highly entangled spin states on a faster timescale than that of the commonly employed one-axis twisting (OAT) model. We demonstrate that by periodically driving the Dicke Hamiltonian at a resonance frequency, the system effectively becomes a two-axis countertwisting Hamiltonian which is known to quickly create Heisenberg limit scaled entangled states. For these states we show that simple quadrature measurements can saturate the ultimate precision limit for parameter estimation determined by the quantum Cramér-Rao bound. An example experimental realization of the periodically driven scheme is discussed with the potential to quickly generate momentum entanglement in a recently described experimental vertical cavity system. We analyze effects of collective dissipation in this vertical cavity system and find that our squeezing protocol can be more robust than the previous realization of OAT.
△ Less
Submitted 12 April, 2024; v1 submitted 11 October, 2023;
originally announced October 2023.
-
Optimal Generators for Quantum Sensing
Authors:
Jarrod T. Reilly,
John Drew Wilson,
Simon B. Jäger,
Christopher Wilson,
Murray J. Holland
Abstract:
We propose a computationally efficient method to derive the unitary evolution that a quantum state is most sensitive to. This allows one to determine the optimal use of an entangled state for quantum sensing, even in complex systems where intuition from canonical squeezing examples breaks down. In this paper we show that the maximal obtainable sensitivity using a given quantum state is determined…
▽ More
We propose a computationally efficient method to derive the unitary evolution that a quantum state is most sensitive to. This allows one to determine the optimal use of an entangled state for quantum sensing, even in complex systems where intuition from canonical squeezing examples breaks down. In this paper we show that the maximal obtainable sensitivity using a given quantum state is determined by the largest eigenvalue of the quantum Fisher information matrix (QFIM) and, importantly, the corresponding evolution is uniquely determined by the coinciding eigenvector. Since we optimize the process of parameter encoding rather than focusing on state preparation protocols, our scheme is relevant for any quantum sensor. This procedure naturally optimizes multiparameter estimation by determining, through the eigenvectors of the QFIM, the maximal set of commuting observables with optimal sensitivity.
△ Less
Submitted 11 August, 2023; v1 submitted 24 May, 2023;
originally announced May 2023.
-
Cavity-Mediated Collective Momentum-Exchange Interactions
Authors:
Chengyi Luo,
Haoqing Zhang,
Vanessa P. W. Koh,
John D. Wilson,
Anjun Chu,
Murray J. Holland,
Ana Maria Rey,
James K. Thompson
Abstract:
Quantum simulation and sensing hold great promise for providing new insights into nature, from understanding complex interacting systems to searching for undiscovered physics. Large ensembles of laser-cooled atoms interacting via infinite-range photon mediated interactions are a powerful platform for both endeavours. Here, we realize for the first time momentum-exchange interactions in which atoms…
▽ More
Quantum simulation and sensing hold great promise for providing new insights into nature, from understanding complex interacting systems to searching for undiscovered physics. Large ensembles of laser-cooled atoms interacting via infinite-range photon mediated interactions are a powerful platform for both endeavours. Here, we realize for the first time momentum-exchange interactions in which atoms exchange their momentum states via collective emission and absorption of photons from a common cavity mode. The momentum-exchange interaction leads to an observed all-to-all Ising-like interaction in a matter-wave interferometer, which is useful for entanglement generation. A many-body energy gap also emerges, effectively binding interferometer matter-wave packets together to suppress Doppler dephasing, akin to Mössbauer spectroscopy. The tunable momentum-exchange interaction provides a new capability for quantum interaction-enhanced matter-wave interferometry and for realizing exotic behaviors including simulations of superconductors and dynamical gauge fields.
△ Less
Submitted 3 April, 2023;
originally announced April 2023.
-
Robust variance-regularized risk minimization with concomitant scaling
Authors:
Matthew J. Holland
Abstract:
Under losses which are potentially heavy-tailed, we consider the task of minimizing sums of the loss mean and standard deviation, without trying to accurately estimate the variance. By modifying a technique for variance-free robust mean estimation to fit our problem setting, we derive a simple learning procedure which can be easily combined with standard gradient-based solvers to be used in tradit…
▽ More
Under losses which are potentially heavy-tailed, we consider the task of minimizing sums of the loss mean and standard deviation, without trying to accurately estimate the variance. By modifying a technique for variance-free robust mean estimation to fit our problem setting, we derive a simple learning procedure which can be easily combined with standard gradient-based solvers to be used in traditional machine learning workflows. Empirically, we verify that our proposed approach, despite its simplicity, performs as well or better than even the best-performing candidates derived from alternative criteria such as CVaR or DRO risks on a variety of datasets.
△ Less
Submitted 8 February, 2024; v1 submitted 27 January, 2023;
originally announced January 2023.
-
Beyond one-axis twisting: Simultaneous spin-momentum squeezing
Authors:
John Drew Wilson,
Simon B. Jäger,
Jarrod T. Reilly,
Athreya Shankar,
Maria Luisa Chiofalo,
Murray J. Holland
Abstract:
The creation and manipulation of quantum entanglement is central to improving precision measurements. A principal method of generating entanglement for use in atom interferometry is the process of spin squeezing whereupon the states become more sensitive to $SU(2)$ rotations. One possibility to generate this entanglement is provided by one-axis twisting (OAT), where a many-particle entangled state…
▽ More
The creation and manipulation of quantum entanglement is central to improving precision measurements. A principal method of generating entanglement for use in atom interferometry is the process of spin squeezing whereupon the states become more sensitive to $SU(2)$ rotations. One possibility to generate this entanglement is provided by one-axis twisting (OAT), where a many-particle entangled state of one degree of freedom is generated by a non-linear Hamiltonian. We introduce a novel method which goes beyond OAT to create squeezing and entanglement across two distinct degrees of freedom. We present our work in the specific physical context of a system consisting of collective atomic energy levels and discrete collective momentum states, but also consider other possible realizations. Our system uses a nonlinear Hamiltonian to generate dynamics in $SU(4)$, thereby creating the opportunity for dynamics not possible in typical $SU(2)$ one-axis twisting. This leads to three axes undergoing twisting due to the two degrees of freedom and their entanglement, with the resulting potential for a more rich context of quantum entanglement. The states prepared in this system are potentially more versatile for use in multi-parameter or auxiliary measurement schemes than those prepared by standard spin squeezing.
△ Less
Submitted 14 September, 2022; v1 submitted 24 June, 2022;
originally announced June 2022.
-
Mean-field Floquet theory for a three-level cold-atom laser
Authors:
Gage W. Harmon,
Jarrod T. Reilly,
Murray J. Holland,
Simon B. Jäger
Abstract:
We present a theoretical description for a lasing scheme for atoms with three internal levels in a $V$-configuration and interacting with an optical cavity. The use of a $V$-level system allows for an efficient closed lasing cycle to be sustained on a dipole-forbidden transition without the need for incoherent repumping. This is made possible by utilizing an additional dipole-allowed transition. W…
▽ More
We present a theoretical description for a lasing scheme for atoms with three internal levels in a $V$-configuration and interacting with an optical cavity. The use of a $V$-level system allows for an efficient closed lasing cycle to be sustained on a dipole-forbidden transition without the need for incoherent repumping. This is made possible by utilizing an additional dipole-allowed transition. We determine the lasing threshold and emission frequency by performing a stability analysis of the non-lasing solution. In the lasing regime, we use a mean-field Floquet method (MFFM) to calculate the lasing intensity and emission frequency. This MFFM predicts the lasing transition to be accompanied by the breaking of a continuous $U(1)$ symmetry in a single Fourier component of the total field. In addition, we use the MFFM to derive bistable lasing and non-lasing solutions that highlight the non-linear nature of this system. We then test the bistability by studying hysteresis when slowly ramping external parameters across the threshold and back. Furthermore, we also compare our mean-field results to a second-order cumulant approach. The work provides simple methods for understanding complex physics that occur in cold atom lasers with narrow line transitions.
△ Less
Submitted 9 May, 2022;
originally announced May 2022.
-
Flexible risk design using bi-directional dispersion
Authors:
Matthew J. Holland
Abstract:
Many novel notions of "risk" (e.g., CVaR, tilted risk, DRO risk) have been proposed and studied, but these risks are all at least as sensitive as the mean to loss tails on the upside, and tend to ignore deviations on the downside. We study a complementary new risk class that penalizes loss deviations in a bi-directional manner, while having more flexibility in terms of tail sensitivity than is off…
▽ More
Many novel notions of "risk" (e.g., CVaR, tilted risk, DRO risk) have been proposed and studied, but these risks are all at least as sensitive as the mean to loss tails on the upside, and tend to ignore deviations on the downside. We study a complementary new risk class that penalizes loss deviations in a bi-directional manner, while having more flexibility in terms of tail sensitivity than is offered by mean-variance. This class lets us derive high-probability learning guarantees without explicit gradient clipping, and empirical tests using both simulated and real data illustrate a high degree of control over key properties of the test loss distribution incurred by gradient-based learners.
△ Less
Submitted 16 February, 2023; v1 submitted 27 March, 2022;
originally announced March 2022.
-
Lindblad master equations for quantum systems coupled to dissipative bosonic modes
Authors:
Simon B. Jäger,
Tom Schmit,
Giovanna Morigi,
Murray J. Holland,
Ralf Betzholz
Abstract:
We present a general approach to derive Lindblad master equations for a subsystem whose dynamics is coupled to dissipative bosonic modes. The derivation relies on a Schrieffer-Wolff transformation which allows to eliminate the bosonic degrees of freedom after self-consistently determining their state as a function of the coupled quantum system. We apply this formalism to the dissipative Dicke mode…
▽ More
We present a general approach to derive Lindblad master equations for a subsystem whose dynamics is coupled to dissipative bosonic modes. The derivation relies on a Schrieffer-Wolff transformation which allows to eliminate the bosonic degrees of freedom after self-consistently determining their state as a function of the coupled quantum system. We apply this formalism to the dissipative Dicke model and derive a Lindblad master equation for the atomic spins, which includes the coherent and dissipative interactions mediated by the bosonic mode. This master equation accurately predicts the Dicke phase transition and gives the correct steady state. In addition, we compare the dynamics using exact diagonalization and numerical integration of the master equation with the predictions of semiclassical trajectories. We finally test the performance of our formalism by studying the relaxation of a NOON state and show that the dynamics captures quantum metastability beyond the mean-field approximation.
△ Less
Submitted 7 March, 2022;
originally announced March 2022.
-
Adiabatic Control of Decoherence-Free-Subspaces in an Open Collective System
Authors:
Jarrod T. Reilly,
Simon B. Jäger,
John Cooper,
Murray J. Holland
Abstract:
We propose a method to adiabatically control an atomic ensemble using a decoherence-free subspace (DFS) within a dissipative cavity. We can engineer a specific eigenstate of the system's Lindblad jump operators by injecting a field into the cavity which deconstructively interferes with the emission amplitude of the ensemble. In contrast to previous adiabatic DFS proposals, our scheme creates a DFS…
▽ More
We propose a method to adiabatically control an atomic ensemble using a decoherence-free subspace (DFS) within a dissipative cavity. We can engineer a specific eigenstate of the system's Lindblad jump operators by injecting a field into the cavity which deconstructively interferes with the emission amplitude of the ensemble. In contrast to previous adiabatic DFS proposals, our scheme creates a DFS in the presence of collective decoherence. We therefore have the ability to engineer states that have high multi-particle entanglements which may be exploited for quantum information science or metrology. We further demonstrate a more optimized driving scheme that utilizes the knowledge of possible diabatic evolution gained from the so-called adiabatic criteria. This allows us to evolve to a desired state with exceptionally high fidelity on a time scale that does not depend on the number of atoms in the ensemble. By engineering the DFS eigenstate adiabatically, our method allows for faster state preparation than previous schemes that rely on damping into a desired state solely using dissipation.
△ Less
Submitted 28 January, 2022;
originally announced January 2022.
-
A Survey of Learning Criteria Going Beyond the Usual Risk
Authors:
Matthew J. Holland,
Kazuki Tanabe
Abstract:
Virtually all machine learning tasks are characterized using some form of loss function, and "good performance" is typically stated in terms of a sufficiently small average loss, taken over the random draw of test data. While optimizing for performance on average is intuitive, convenient to analyze in theory, and easy to implement in practice, such a choice brings about trade-offs. In this work, w…
▽ More
Virtually all machine learning tasks are characterized using some form of loss function, and "good performance" is typically stated in terms of a sufficiently small average loss, taken over the random draw of test data. While optimizing for performance on average is intuitive, convenient to analyze in theory, and easy to implement in practice, such a choice brings about trade-offs. In this work, we survey and introduce a wide variety of non-traditional criteria used to design and evaluate machine learning algorithms, place the classical paradigm within the proper historical context, and propose a view of learning problems which emphasizes the question of "what makes for a desirable loss distribution?" in place of tacit use of the expected loss.
△ Less
Submitted 29 November, 2023; v1 submitted 11 October, 2021;
originally announced October 2021.
-
Collective emission of an atomic beam into an off-resonant cavity mode
Authors:
Simon B. Jäger,
Haonan Liu,
John Cooper,
Murray J. Holland
Abstract:
We study the collective emission of a beam of atomic dipoles into an optical cavity. Our focus lies on the effect of a finite detuning between the atomic transition frequency and the cavity resonance frequency. By developing a theoretical description of the coupled atom-cavity dynamics we analyze the stationary atomic configurations including a superradiant phase where the atoms undergo continuous…
▽ More
We study the collective emission of a beam of atomic dipoles into an optical cavity. Our focus lies on the effect of a finite detuning between the atomic transition frequency and the cavity resonance frequency. By developing a theoretical description of the coupled atom-cavity dynamics we analyze the stationary atomic configurations including a superradiant phase where the atoms undergo continuous monochromatic collective emission. In addition, we derive an analytical formula for the cavity pulling coefficient which characterizes the displacement of the emission frequency towards the cavity frequency. We find that the pulling is small if the cavity linewidth is much larger than the collective linewidth of the atomic beam. This regime is desired for building stable lasers because the emission frequency is robust against cavity length fluctuations. Furthermore, we investigate the stability of the atomic phases and compare our theoretical predictions with numerical results. Remarkably, we also find polychromatic emission regimes, where the spectrum has several frequency components while the light output is still superradiant.
△ Less
Submitted 12 July, 2021;
originally announced July 2021.
-
Robust learning with anytime-guaranteed feedback
Authors:
Matthew J. Holland
Abstract:
Under data distributions which may be heavy-tailed, many stochastic gradient-based learning algorithms are driven by feedback queried at points with almost no performance guarantees on their own. Here we explore a modified "anytime online-to-batch" mechanism which for smooth objectives admits high-probability error bounds while requiring only lower-order moment bounds on the stochastic gradients.…
▽ More
Under data distributions which may be heavy-tailed, many stochastic gradient-based learning algorithms are driven by feedback queried at points with almost no performance guarantees on their own. Here we explore a modified "anytime online-to-batch" mechanism which for smooth objectives admits high-probability error bounds while requiring only lower-order moment bounds on the stochastic gradients. Using this conversion, we can derive a wide variety of "anytime robust" procedures, for which the task of performance analysis can be effectively reduced to regret control, meaning that existing regret bounds (for the bounded gradient case) can be robustified and leveraged in a straightforward manner. As a direct takeaway, we obtain an easily implemented stochastic gradient-based algorithm for which all queried points formally enjoy sub-Gaussian error bounds, and in practice show noteworthy gains on real-world data applications.
△ Less
Submitted 24 May, 2021;
originally announced May 2021.
-
Superradiant emission of a thermal atomic beam into an optical cavity
Authors:
Simon B. Jäger,
Haonan Liu,
John Cooper,
Travis L. Nicholson,
Murray J. Holland
Abstract:
We theoretically analyze the collective dynamics of a thermal beam of atomic dipoles that couple to a single mode when traversing an optical cavity. For this setup we derive a semiclassical model and determine the onset of superradiant emission and its stability. We derive analytical expressions for the linewidth of the emitted light and compare them with numerical simulations. In addition, we fin…
▽ More
We theoretically analyze the collective dynamics of a thermal beam of atomic dipoles that couple to a single mode when traversing an optical cavity. For this setup we derive a semiclassical model and determine the onset of superradiant emission and its stability. We derive analytical expressions for the linewidth of the emitted light and compare them with numerical simulations. In addition, we find and predict two different superradiant phases; a steady-state superradiant phase and a multi-component superradiant phase. In the latter case we observe sidebands in the frequency spectrum that can be calculated using a stability analysis of the amplitude mode of the collective dipole. We show that both superradiant phases are robust against free-space spontaneous emission and $T_2$ dephasing processes.
△ Less
Submitted 18 May, 2021;
originally announced May 2021.
-
Spectral risk-based learning using unbounded losses
Authors:
Matthew J. Holland,
El Mehdi Haress
Abstract:
In this work, we consider the setting of learning problems under a wide class of spectral risk (or "L-risk") functions, where a Lipschitz-continuous spectral density is used to flexibly assign weight to extreme loss values. We obtain excess risk guarantees for a derivative-free learning procedure under unbounded heavy-tailed loss distributions, and propose a computationally efficient implementatio…
▽ More
In this work, we consider the setting of learning problems under a wide class of spectral risk (or "L-risk") functions, where a Lipschitz-continuous spectral density is used to flexibly assign weight to extreme loss values. We obtain excess risk guarantees for a derivative-free learning procedure under unbounded heavy-tailed loss distributions, and propose a computationally efficient implementation which empirically outperforms traditional risk minimizers in terms of balancing spectral risk and misclassification error.
△ Less
Submitted 11 May, 2021;
originally announced May 2021.
-
Entropy transfer from a quantum particle to a classical coherent light field
Authors:
John P. Bartolotta,
Simon B. Jäger,
Jarrod T. Reilly,
Matthew A. Norcia,
James K. Thompson,
Graeme Smith,
Murray J. Holland
Abstract:
In the field of light-matter interactions, it is often assumed that a classical light field that interacts with a quantum particle remains almost unchanged and thus contains nearly no information about the manipulated particles. To investigate the validity of this assumption, we develop and theoretically analyze a simple Gedankenexperiment which involves the interaction of a coherent state with a…
▽ More
In the field of light-matter interactions, it is often assumed that a classical light field that interacts with a quantum particle remains almost unchanged and thus contains nearly no information about the manipulated particles. To investigate the validity of this assumption, we develop and theoretically analyze a simple Gedankenexperiment which involves the interaction of a coherent state with a quantum particle in an optical cavity. We quantify the resulting alteration of the light field by measuring the fidelity of its initial and equilibrium states. Using Bayesian inference, we demonstrate the information transfer through photon measurements. In addition, we employ the concepts of quantum entropy and mutual information to quantify the entropy transfer from the particle to the light field. In the weak coupling limit, we validate the usually assumed negligible alteration of the light field and entropy transfer. In the strong coupling limit, however, we observe that the information of the initial particle state can be fully encoded in the light field, even for large photon numbers. Nevertheless, we show that spontaneous emission is a sufficient mechanism for removing the entropy initially stored in the particle. Our analysis provides a deeper understanding of the entropy exchange between quantum matter and classical light.
△ Less
Submitted 8 May, 2021;
originally announced May 2021.
-
Subradiant-to-Subradiant Phase Transition in the Bad Cavity Laser
Authors:
Athreya Shankar,
Jarrod T. Reilly,
Simon B. Jäger,
Murray J. Holland
Abstract:
We show that the onset of steady-state superradiance in a bad cavity laser is preceded by a dissipative phase transition between two distinct phases of steady-state subradiance. The transition is marked by a non-analytic behavior of the cavity output power and the mean atomic inversion, as well as a discontinuity in the variance of the collective atomic inversion. In particular, for repump rates b…
▽ More
We show that the onset of steady-state superradiance in a bad cavity laser is preceded by a dissipative phase transition between two distinct phases of steady-state subradiance. The transition is marked by a non-analytic behavior of the cavity output power and the mean atomic inversion, as well as a discontinuity in the variance of the collective atomic inversion. In particular, for repump rates below a critical value, the cavity output power is strongly suppressed and does not increase with the atom number, while it scales linearly with atom number above this value. Remarkably, we find that the atoms are in a macroscopic entangled steady state near the critical region with a vanishing fraction of unentangled atoms in the large atom number limit.
△ Less
Submitted 12 March, 2021;
originally announced March 2021.
-
Better scalability under potentially heavy-tailed feedback
Authors:
Matthew J. Holland
Abstract:
We study scalable alternatives to robust gradient descent (RGD) techniques that can be used when the losses and/or gradients can be heavy-tailed, though this will be unknown to the learner. The core technique is simple: instead of trying to robustly aggregate gradients at each step, which is costly and leads to sub-optimal dimension dependence in risk bounds, we instead focus computational effort…
▽ More
We study scalable alternatives to robust gradient descent (RGD) techniques that can be used when the losses and/or gradients can be heavy-tailed, though this will be unknown to the learner. The core technique is simple: instead of trying to robustly aggregate gradients at each step, which is costly and leads to sub-optimal dimension dependence in risk bounds, we instead focus computational effort on robustly choosing (or newly constructing) a strong candidate based on a collection of cheap stochastic sub-processes which can be run in parallel. The exact selection process depends on the convexity of the underlying objective, but in all cases, our selection technique amounts to a robust form of boosting the confidence of weak learners. In addition to formal guarantees, we also provide empirical analysis of robustness to perturbations to experimental conditions, under both sub-Gaussian and heavy-tailed data, along with applications to a variety of benchmark datasets. The overall take-away is an extensible procedure that is simple to implement, trivial to parallelize, which keeps the formal merits of RGD methods but scales much better to large learning problems.
△ Less
Submitted 14 December, 2020;
originally announced December 2020.
-
Learning with risks based on M-location
Authors:
Matthew J. Holland
Abstract:
In this work, we study a new class of risks defined in terms of the location and deviation of the loss distribution, generalizing far beyond classical mean-variance risk functions. The class is easily implemented as a wrapper around any smooth loss, it admits finite-sample stationarity guarantees for stochastic gradient methods, it is straightforward to interpret and adjust, with close links to M-…
▽ More
In this work, we study a new class of risks defined in terms of the location and deviation of the loss distribution, generalizing far beyond classical mean-variance risk functions. The class is easily implemented as a wrapper around any smooth loss, it admits finite-sample stationarity guarantees for stochastic gradient methods, it is straightforward to interpret and adjust, with close links to M-estimators of the loss location, and has a salient effect on the test loss distribution.
△ Less
Submitted 25 April, 2021; v1 submitted 4 December, 2020;
originally announced December 2020.
-
Rugged mHz-Linewidth Superradiant Laser Driven by a Hot Atomic Beam
Authors:
Haonan Liu,
Simon B. Jäger,
Xianquan Yu,
Steven Touzard,
Athreya Shankar,
Murray J. Holland,
Travis L. Nicholson
Abstract:
We propose a new type of superradiant laser based on a hot atomic beam traversing an optical cavity. We show that the theoretical minimum linewidth and maximum power are competitive with the best ultracoherent clock lasers. Also, our system operates naturally in continuous wave mode, which has been elusive for superradiant lasers so far. Unlike existing ultracoherent lasers, our design is simple a…
▽ More
We propose a new type of superradiant laser based on a hot atomic beam traversing an optical cavity. We show that the theoretical minimum linewidth and maximum power are competitive with the best ultracoherent clock lasers. Also, our system operates naturally in continuous wave mode, which has been elusive for superradiant lasers so far. Unlike existing ultracoherent lasers, our design is simple and rugged. This makes it a candidate for the first widely accessible ultracoherent laser, as well as the first to realize sought-after applications of ultracoherent lasers in challenging environments.
△ Less
Submitted 23 December, 2020; v1 submitted 11 September, 2020;
originally announced September 2020.
-
Regular and bistable steady-state superradiant phases of an atomic beam traversing an optical cavity
Authors:
Simon B. Jäger,
Haonan Liu,
Athreya Shankar,
John Cooper,
Murray J. Holland
Abstract:
We investigate the different photon emission regimes created by a preexcited and collimated atomic beam passing through a single mode of an optical cavity. In the regime where the cavity degrees of freedom can be adiabatically eliminated, we find that the atoms undergo superradiant emission when the collective linewidth exceeds the transit-time broadening. We analyze the case where the atomic beam…
▽ More
We investigate the different photon emission regimes created by a preexcited and collimated atomic beam passing through a single mode of an optical cavity. In the regime where the cavity degrees of freedom can be adiabatically eliminated, we find that the atoms undergo superradiant emission when the collective linewidth exceeds the transit-time broadening. We analyze the case where the atomic beam direction is slanted with respect to the cavity axis. For this situation, we find that a phase of continuous light emission similar to steady-state superradiance is established providing the tilt of the atomic beam is sufficiently small. However, if the atoms travel more than half a wavelength along the cavity axis during one transit time we predict a dynamical phase transition to a new bistable superradiant regime. In this phase the atoms undergo collective spontaneous emission with a frequency that can be either blue or red detuned from the free-space atomic resonance. We analyze the different superradiant regimes and the quantum critical crossover boundaries. In particular we find the spectrum of the emitted light and show that the linewidth exhibits features of a critical scaling close to the phase boundaries.
△ Less
Submitted 13 November, 2020; v1 submitted 11 September, 2020;
originally announced September 2020.
-
Broadening of the drumhead mode spectrum due to in-plane thermal fluctuations of two-dimensional trapped ion crystals in a Penning trap
Authors:
Athreya Shankar,
Chen Tang,
Matthew Affolter,
Kevin Gilmore,
Daniel H. E. Dubin,
Scott Parker,
Murray J. Holland,
John J. Bollinger
Abstract:
Two-dimensional crystals of ions stored in Penning traps are a leading platform for quantum simulation and sensing experiments. For small amplitudes, the out-of-plane motion of such crystals can be described by a discrete set of normal modes called the drumhead modes, which can be used to implement a range of quantum information protocols. However, experimental observations of crystals with Dopple…
▽ More
Two-dimensional crystals of ions stored in Penning traps are a leading platform for quantum simulation and sensing experiments. For small amplitudes, the out-of-plane motion of such crystals can be described by a discrete set of normal modes called the drumhead modes, which can be used to implement a range of quantum information protocols. However, experimental observations of crystals with Doppler-cooled and even near-ground-state-cooled drumhead modes reveal an unresolved drumhead mode spectrum. In this work, we establish in-plane thermal fluctuations in ion positions as a major contributor to the broadening of the drumhead mode spectrum. In the process, we demonstrate how the confining magnetic field leads to unconventional in-plane normal modes, whose average potential and kinetic energies are not equal. This property, in turn, has implications for the sampling procedure required to choose the in-plane initial conditions for molecular dynamics simulations. For current operating conditions of the NIST Penning trap, our study suggests that the two dimensional crystals produced in this trap undergo in-plane potential energy fluctuations of the order of $10$ mK. Our study therefore motivates the need for designing improved techniques to cool the in-plane degrees of freedom.
△ Less
Submitted 18 August, 2020;
originally announced August 2020.
-
Speeding Up Particle Slowing using Shortcuts to Adiabaticity
Authors:
John P. Bartolotta,
Jarrod T. Reilly,
Murray J. Holland
Abstract:
We propose a method for slowing particles by laser fields that potentially has the ability to generate large forces without the associated momentum diffusion that results from the random directions of spontaneously scattered photons. In this method, time-resolved laser pulses with periodically modified detunings address an ultranarrow electronic transition to reduce the particle momentum through r…
▽ More
We propose a method for slowing particles by laser fields that potentially has the ability to generate large forces without the associated momentum diffusion that results from the random directions of spontaneously scattered photons. In this method, time-resolved laser pulses with periodically modified detunings address an ultranarrow electronic transition to reduce the particle momentum through repeated absorption and stimulated emission cycles. We implement a shortcut to adiabaticity approach that is based on Lewis-Riesenfeld invariant theory. This affords our scheme the advantages of adiabatic transfer, where there can be an intrinsic insensitivity to the precise strength and detuning characteristics of the applied field, with the advantages of rapid transfer that is necessary for obtaining a short slowing distance. For typical parameters of a thermal oven source that generates a particle beam with a central velocity on the order of meters per second, this could result in slowing the particles to near stationary in less than a millimeter. We compare the slowing scheme to widely-implemented slowing techniques that rely on radiation pressure forces and show the advantages that potentially arise when the excited state decay rate is small. Thus, this scheme is a particularly promising candidate to slow narrow-linewidth systems that lack closed cycling transitions, such as occurs in certain molecules.
△ Less
Submitted 13 July, 2020;
originally announced July 2020.
-
Making learning more transparent using conformalized performance prediction
Authors:
Matthew J. Holland
Abstract:
In this work, we study some novel applications of conformal inference techniques to the problem of providing machine learning procedures with more transparent, accurate, and practical performance guarantees. We provide a natural extension of the traditional conformal prediction framework, done in such a way that we can make valid and well-calibrated predictive statements about the future performan…
▽ More
In this work, we study some novel applications of conformal inference techniques to the problem of providing machine learning procedures with more transparent, accurate, and practical performance guarantees. We provide a natural extension of the traditional conformal prediction framework, done in such a way that we can make valid and well-calibrated predictive statements about the future performance of arbitrary learning algorithms, when passed an as-yet unseen training set. In addition, we include some nascent empirical examples to illustrate potential applications.
△ Less
Submitted 8 July, 2020;
originally announced July 2020.
-
Learning with CVaR-based feedback under potentially heavy tails
Authors:
Matthew J. Holland,
El Mehdi Haress
Abstract:
We study learning algorithms that seek to minimize the conditional value-at-risk (CVaR), when all the learner knows is that the losses incurred may be heavy-tailed. We begin by studying a general-purpose estimator of CVaR for potentially heavy-tailed random variables, which is easy to implement in practice, and requires nothing more than finite variance and a distribution function that does not ch…
▽ More
We study learning algorithms that seek to minimize the conditional value-at-risk (CVaR), when all the learner knows is that the losses incurred may be heavy-tailed. We begin by studying a general-purpose estimator of CVaR for potentially heavy-tailed random variables, which is easy to implement in practice, and requires nothing more than finite variance and a distribution function that does not change too fast or slow around just the quantile of interest. With this estimator in hand, we then derive a new learning algorithm which robustly chooses among candidates produced by stochastic gradient-driven sub-processes. For this procedure we provide high-probability excess CVaR bounds, and to complement the theory we conduct empirical tests of the underlying CVaR estimator and the learning algorithm derived from it.
△ Less
Submitted 2 June, 2020;
originally announced June 2020.
-
Improved scalability under heavy tails, without strong convexity
Authors:
Matthew J. Holland
Abstract:
Real-world data is laden with outlying values. The challenge for machine learning is that the learner typically has no prior knowledge of whether the feedback it receives (losses, gradients, etc.) will be heavy-tailed or not. In this work, we study a simple algorithmic strategy that can be leveraged when both losses and gradients can be heavy-tailed. The core technique introduces a simple robust v…
▽ More
Real-world data is laden with outlying values. The challenge for machine learning is that the learner typically has no prior knowledge of whether the feedback it receives (losses, gradients, etc.) will be heavy-tailed or not. In this work, we study a simple algorithmic strategy that can be leveraged when both losses and gradients can be heavy-tailed. The core technique introduces a simple robust validation sub-routine, which is used to boost the confidence of inexpensive gradient-based sub-processes. Compared with recent robust gradient descent methods from the literature, dimension dependence (both risk bounds and cost) is substantially improved, without relying upon strong convexity or expensive per-step robustification. Empirically, we also show that under heavy-tailed losses, the proposed procedure cannot simply be replaced with naive cross-validation. Taken together, we have a scalable method with transparent guarantees, which performs well without prior knowledge of how "convenient" the feedback it receives will be.
△ Less
Submitted 14 December, 2020; v1 submitted 1 June, 2020;
originally announced June 2020.
-
Better scalability under potentially heavy-tailed gradients
Authors:
Matthew J. Holland
Abstract:
We study a scalable alternative to robust gradient descent (RGD) techniques that can be used when the gradients can be heavy-tailed, though this will be unknown to the learner. The core technique is simple: instead of trying to robustly aggregate gradients at each step, which is costly and leads to sub-optimal dimension dependence in risk bounds, we choose a candidate which does not diverge too fa…
▽ More
We study a scalable alternative to robust gradient descent (RGD) techniques that can be used when the gradients can be heavy-tailed, though this will be unknown to the learner. The core technique is simple: instead of trying to robustly aggregate gradients at each step, which is costly and leads to sub-optimal dimension dependence in risk bounds, we choose a candidate which does not diverge too far from the majority of cheap stochastic sub-processes run for a single pass over partitioned data. In addition to formal guarantees, we also provide empirical analysis of robustness to perturbations to experimental conditions, under both sub-Gaussian and heavy-tailed data. The result is a procedure that is simple to implement, trivial to parallelize, which keeps the formal strength of RGD methods but scales much better to large learning problems.
△ Less
Submitted 14 December, 2020; v1 submitted 1 June, 2020;
originally announced June 2020.
-
Sawtooth Wave Adiabatic Passage in a Magneto-Optical Trap
Authors:
John P. Bartolotta,
Murray J. Holland
Abstract:
We investigate theoretically the application of Sawtooth Wave Adiabatic Passage (SWAP) in a 1D magneto-optical trap (MOT). As opposed to related methods that have been previously discussed, our approach utilizes repeated cycles of stimulated absorption and emission processes to achieve both trapping and cooling, thereby reducing the adverse effects that arise from photon scattering. Specifically,…
▽ More
We investigate theoretically the application of Sawtooth Wave Adiabatic Passage (SWAP) in a 1D magneto-optical trap (MOT). As opposed to related methods that have been previously discussed, our approach utilizes repeated cycles of stimulated absorption and emission processes to achieve both trapping and cooling, thereby reducing the adverse effects that arise from photon scattering. Specifically, we demonstrate this method's ability to cool, slow, and trap particles with fewer spontaneously emitted photons, higher forces and in less time when compared to a traditional MOT scheme that utilizes the same narrow linewidth transition. We calculate the phase space compression that is achievable and characterize the resulting system equilibrium cloud size and temperature.
△ Less
Submitted 14 May, 2020; v1 submitted 21 January, 2020;
originally announced January 2020.
-
Superradiant optomechanical phases of cold atomic gases in optical resonators
Authors:
Simon B. Jäger,
Murray J. Holland,
Giovanna Morigi
Abstract:
We theoretically analyze superradiant emission of light from a cold atomic gas, when mechanical effects of photon-atom interactions are considered. The atoms are confined within a standing-wave resonator and an atomic metastable dipolar transition couples to a cavity mode. The atomic dipole is incoherently pumped in the parameter regime that would correspond to stationary superradiance in absence…
▽ More
We theoretically analyze superradiant emission of light from a cold atomic gas, when mechanical effects of photon-atom interactions are considered. The atoms are confined within a standing-wave resonator and an atomic metastable dipolar transition couples to a cavity mode. The atomic dipole is incoherently pumped in the parameter regime that would correspond to stationary superradiance in absence of inhomogeneous broadening. Starting from the master equation for cavity field and atomic degrees of freedom we derive a mean-field model that allows us to determine a threshold temperature, above which thermal fluctuations suppress superradiant emission. We then analyze the dynamics of superradiant emission when the motion is described by a mean-field model. In the semiclassical regime and below the threshold temperature we observe that the emitted light can be either coherent or chaotic, depending on the incoherent pump rate. We then analyze superradiant emission from an ideal Bose gas at zero temperature when the superradiant decay rate $Λ$ is of the order of the recoil frequency $ω_R$. We show that the quantized exchange of mechanical energy between the atoms and the field gives rise to a threshold, $Λ_c$, below which superradiant emission is damped down to zero. When $Λ>Λ_c$ superradiant emission is accompanied by the formation of matter-wave gratings diffracting the emitted photons. The stability of these gratings depends on the incoherent pump rate $w$ with respect to a second threshold value $w_c$. For $w>w_c$ the gratings are stable and the system achieves stationary superradiance. Below this second threshold the coupled dynamics becomes chaotic. We characterize the dynamics across these two thresholds and show that the three phases we predict (incoherent, coherent, chaotic) can be revealed via the coherence properties of the light at the cavity output.
△ Less
Submitted 6 November, 2019;
originally announced November 2019.
-
Squeezed state metrology with Bragg interferometers operating in a cavity
Authors:
Athreya Shankar,
Leonardo Salvi,
Maria Luisa Chiofalo,
Nicola Poli,
Murray J Holland
Abstract:
Bragg interferometers, operating using pseudospin-1/2 systems composed of two momentum states, have become a mature technology for precision measurements. State-of-the-art Bragg interferometers are rapidly surpassing technical limitations and are soon expected to operate near the projection noise limit set by uncorrelated atoms. Despite the use of large numbers of atoms, their operation is governe…
▽ More
Bragg interferometers, operating using pseudospin-1/2 systems composed of two momentum states, have become a mature technology for precision measurements. State-of-the-art Bragg interferometers are rapidly surpassing technical limitations and are soon expected to operate near the projection noise limit set by uncorrelated atoms. Despite the use of large numbers of atoms, their operation is governed by single-atom physics. Motivated by recent proposals and demonstrations of Raman gravimeters in cavities, we propose a scheme to squeeze directly on momentum states for surpassing the projection noise limit in Bragg interferometers. In our modeling, we consider the unique issues that arise when a spin squeezing protocol is applied to momentum pseudospins. Specifically, we study the effects of the momentum width of the atomic cloud and the coupling to momentum states outside the pseudospin manifold, as these atoms interact via a mode of the cavity. We show that appreciable levels of spin squeezing can be demonstrated in suitable parameter regimes in spite of these complications. Using this setting, we show how beyond mean-field techniques developed for spin systems can be adapted to study the dynamics of momentum states of interacting atoms. Our scheme promises to be feasible using current technology and is experimentally attractive because it requires no additional setup beyond what will be required to operate Bragg interferometers in cavities.
△ Less
Submitted 23 July, 2019;
originally announced July 2019.
-
Distribution-robust mean estimation via smoothed random perturbations
Authors:
Matthew J. Holland
Abstract:
We consider the problem of mean estimation assuming only finite variance. We study a new class of mean estimators constructed by integrating over random noise applied to a soft-truncated empirical mean estimator. For appropriate choices of noise, we show that this can be computed in closed form, and utilizing relative entropy inequalities, these estimators enjoy deviations with exponential tails c…
▽ More
We consider the problem of mean estimation assuming only finite variance. We study a new class of mean estimators constructed by integrating over random noise applied to a soft-truncated empirical mean estimator. For appropriate choices of noise, we show that this can be computed in closed form, and utilizing relative entropy inequalities, these estimators enjoy deviations with exponential tails controlled by the second moment of the underlying distribution. We consider both additive and multiplicative noise, and several noise distribution families in our analysis. Furthermore, we empirically investigate the sensitivity to the mean-standard deviation ratio for numerous concrete manifestations of the estimator class of interest. Our main take-away is that an inexpensive new estimator can achieve nearly sub-Gaussian performance for a wide variety of data distributions.
△ Less
Submitted 24 June, 2019;
originally announced June 2019.
-
PAC-Bayes under potentially heavy tails
Authors:
Matthew J. Holland
Abstract:
We derive PAC-Bayesian learning guarantees for heavy-tailed losses, and obtain a novel optimal Gibbs posterior which enjoys finite-sample excess risk bounds at logarithmic confidence. Our core technique itself makes use of PAC-Bayesian inequalities in order to derive a robust risk estimator, which by design is easy to compute. In particular, only assuming that the first three moments of the loss d…
▽ More
We derive PAC-Bayesian learning guarantees for heavy-tailed losses, and obtain a novel optimal Gibbs posterior which enjoys finite-sample excess risk bounds at logarithmic confidence. Our core technique itself makes use of PAC-Bayesian inequalities in order to derive a robust risk estimator, which by design is easy to compute. In particular, only assuming that the first three moments of the loss distribution are bounded, the learning algorithm derived from this estimator achieves nearly sub-Gaussian statistical error, up to the quality of the prior.
△ Less
Submitted 18 December, 2019; v1 submitted 20 May, 2019;
originally announced May 2019.
-
Dynamical phase transitions to optomechanical superradiance
Authors:
Simon B. Jäger,
John Cooper,
Murray J. Holland,
Giovanna Morigi
Abstract:
We theoretically analyze superradiant emission of light from an ultracold gas of bosonic atoms confined in a bad cavity. A metastable dipolar transition of the atoms couples to the cavity field and is incoherently pumped, the mechanical effects of cavity-atom interactions tend to order the atoms in the periodic cavity potential. By means of a mean-field model we determine the conditions on the cav…
▽ More
We theoretically analyze superradiant emission of light from an ultracold gas of bosonic atoms confined in a bad cavity. A metastable dipolar transition of the atoms couples to the cavity field and is incoherently pumped, the mechanical effects of cavity-atom interactions tend to order the atoms in the periodic cavity potential. By means of a mean-field model we determine the conditions on the cavity parameters and pump rate that lead to the buildup of a stable macroscopic dipole emitting coherent light. We show that this occurs when the superradiant decay rate and the pump rate exceed threshold values of the order of the photon recoil energy. Above these thresholds superradiant emission is accompanied by the formation of stable matter-wave gratings that diffract the emitted photons. Outside of this regime, instead, the optomechanical coupling can give rise to dephasing or chaos, for which the emitted light is respectively incoherent or chaotic. These behaviors exhibit the features of a dynamical phase transitions and emerge from the interplay between global optomechanical interactions, quantum fluctuations, and noise.
△ Less
Submitted 4 June, 2019; v1 submitted 26 November, 2018;
originally announced November 2018.
-
Robust descent using smoothed multiplicative noise
Authors:
Matthew J. Holland
Abstract:
To improve the off-sample generalization of classical procedures minimizing the empirical risk under potentially heavy-tailed data, new robust learning algorithms have been proposed in recent years, with generalized median-of-means strategies being particularly salient. These procedures enjoy performance guarantees in the form of sharp risk bounds under weak moment assumptions on the underlying lo…
▽ More
To improve the off-sample generalization of classical procedures minimizing the empirical risk under potentially heavy-tailed data, new robust learning algorithms have been proposed in recent years, with generalized median-of-means strategies being particularly salient. These procedures enjoy performance guarantees in the form of sharp risk bounds under weak moment assumptions on the underlying loss, but typically suffer from a large computational overhead and substantial bias when the data happens to be sub-Gaussian, limiting their utility. In this work, we propose a novel robust gradient descent procedure which makes use of a smoothed multiplicative noise applied directly to observations before constructing a sum of soft-truncated gradient coordinates. We show that the procedure has competitive theoretical guarantees, with the major advantage of a simple implementation that does not require an iterative sub-routine for robustification. Empirical tests reinforce the theory, showing more efficient generalization over a much wider class of data distributions.
△ Less
Submitted 15 October, 2018;
originally announced October 2018.
-
Classification using margin pursuit
Authors:
Matthew J. Holland
Abstract:
In this work, we study a new approach to optimizing the margin distribution realized by binary classifiers. The classical approach to this problem is simply maximization of the expected margin, while more recent proposals consider simultaneous variance control and proxy objectives based on robust location estimates, in the vein of keeping the margin distribution sharply concentrated in a desirable…
▽ More
In this work, we study a new approach to optimizing the margin distribution realized by binary classifiers. The classical approach to this problem is simply maximization of the expected margin, while more recent proposals consider simultaneous variance control and proxy objectives based on robust location estimates, in the vein of keeping the margin distribution sharply concentrated in a desirable region. While conceptually appealing, these new approaches are often computationally unwieldy, and theoretical guarantees are limited. Given this context, we propose an algorithm which searches the hypothesis space in such a way that a pre-set "margin level" ends up being a distribution-robust estimator of the margin location. This procedure is easily implemented using gradient descent, and admits finite-sample bounds on the excess risk under unbounded inputs. Empirical tests on real-world benchmark data reinforce the basic principles highlighted by the theory, and are suggestive of a promising new technique for classification.
△ Less
Submitted 11 October, 2018;
originally announced October 2018.
-
Near ground-state cooling of two-dimensional trapped-ion crystals with more than 100 ions
Authors:
Elena Jordan,
Kevin A. Gilmore,
Athreya Shankar,
Arghavan Safavi-Naini,
Justin G. Bohnet,
Murray J. Holland,
John J. Bollinger
Abstract:
We study, both experimentally and theoretically, electromagnetically induced transparency cooling of the drumhead modes of planar 2-dimensional arrays with up to $N\approx 190$ Be${}^+$ ions stored in a Penning trap. Substantial sub-Doppler cooling is observed for all $N$ drumhead modes. Quantitative measurements for the center-of-mass mode show near ground state cooling with motional quantum numb…
▽ More
We study, both experimentally and theoretically, electromagnetically induced transparency cooling of the drumhead modes of planar 2-dimensional arrays with up to $N\approx 190$ Be${}^+$ ions stored in a Penning trap. Substantial sub-Doppler cooling is observed for all $N$ drumhead modes. Quantitative measurements for the center-of-mass mode show near ground state cooling with motional quantum numbers of $\bar{n} = 0.3\pm0.2$ obtained within $200~μs$. The measured cooling rate is faster than that predicted by single particle theory, consistent with a quantum many-body calculation. For the lower frequency drumhead modes, quantitative temperature measurements are limited by apparent damping and frequency instabilities, but near ground state cooling of the full bandwidth is strongly suggested. This advancement will greatly improve the performance of large trapped ion crystals in quantum information and quantum metrology applications.
△ Less
Submitted 17 September, 2018;
originally announced September 2018.
-
Modeling near ground-state cooling of two-dimensional ion crystals in a Penning trap using electromagnetically induced transparency
Authors:
Athreya Shankar,
Elena Jordan,
Kevin A. Gilmore,
Arghavan Safavi-Naini,
John J. Bollinger,
Murray J. Holland
Abstract:
Penning traps, with their ability to control planar crystals of tens to hundreds of ions, are versatile quantum simulators. Thermal occupations of the motional drumhead modes, transverse to the plane of the ion crystal, degrade the quality of quantum simulations. Laser cooling using electromagnetically induced transparency (EIT cooling) is attractive as an efficient way to quickly initialize the d…
▽ More
Penning traps, with their ability to control planar crystals of tens to hundreds of ions, are versatile quantum simulators. Thermal occupations of the motional drumhead modes, transverse to the plane of the ion crystal, degrade the quality of quantum simulations. Laser cooling using electromagnetically induced transparency (EIT cooling) is attractive as an efficient way to quickly initialize the drumhead modes to near ground-state occupations. We numerically investigate the efficiency of EIT cooling of planar ion crystals in a Penning trap, accounting for complications arising from the nature of the trap and from the simultaneous cooling of multiple ions. We show that, in spite of challenges, the large bandwidth of drumhead modes (hundreds of kilohertz) can be rapidly cooled to near ground-state occupations within a few hundred microseconds. Our predictions for the center-of-mass mode include a cooling time constant of tens of microseconds and an enhancement of the cooling rate with increasing number of ions. Successful experimental demonstrations of EIT cooling in the NIST Penning trap [E. Jordan, K. A. Gilmore, A. Shankar, A. Safavi-Naini, M. J. Holland, and J. J. Bollinger, "Near ground-state cooling of two-dimensional trapped-ion crystals with more than 100 ions", (2018), submitted.] validate our predictions.
△ Less
Submitted 18 September, 2018; v1 submitted 14 September, 2018;
originally announced September 2018.
-
Laser Cooling by Sawtooth Wave Adiabatic Passage
Authors:
John P. Bartolotta,
Matthew A. Norcia,
Julia R. K. Cline,
James K. Thompson,
Murray J. Holland
Abstract:
We provide a theoretical analysis for a recently demonstrated cooling method. Two-level particles undergo successive adiabatic transfers upon interaction with counter-propagating laser beams that are repeatedly swept over the transition frequency. We show that particles with narrow linewidth transitions can be cooled to near the recoil limit. This cooling mechanism has a reduced reliance on sponta…
▽ More
We provide a theoretical analysis for a recently demonstrated cooling method. Two-level particles undergo successive adiabatic transfers upon interaction with counter-propagating laser beams that are repeatedly swept over the transition frequency. We show that particles with narrow linewidth transitions can be cooled to near the recoil limit. This cooling mechanism has a reduced reliance on spontaneous emission compared to Doppler cooling, and hence shows promise for application to systems lacking closed cycling transitions, such as molecules.
△ Less
Submitted 7 June, 2018;
originally announced June 2018.
-
Narrow-line Laser Cooling by Adiabatic Transfer
Authors:
Matthew A. Norcia,
Julia R. K. Cline,
John P. Bartolotta,
Murray J. Holland,
James K. Thompson
Abstract:
We propose and demonstrate a novel laser cooling mechanism applicable to particles with narrow-linewidth optical transitions. By sweeping the frequency of counter-propagating laser beams in a sawtooth manner, we cause adiabatic transfer back and forth between the ground state and a long-lived optically excited state. The time-ordering of these adiabatic transfers is determined by Doppler shifts, w…
▽ More
We propose and demonstrate a novel laser cooling mechanism applicable to particles with narrow-linewidth optical transitions. By sweeping the frequency of counter-propagating laser beams in a sawtooth manner, we cause adiabatic transfer back and forth between the ground state and a long-lived optically excited state. The time-ordering of these adiabatic transfers is determined by Doppler shifts, which ensures that the associated photon recoils are in the opposite direction to the particle's motion. This ultimately leads to a robust cooling mechanism capable of exerting large forces via a weak transition and with reduced reliance on spontaneous emission. We present a simple intuitive model for the resulting frictional force, and directly demonstrate its efficacy for increasing the total phase-space density of an atomic ensemble. We rely on both simulation and experimental studies using the 7.5~kHz linewidth $^1$S$_0$ to $^3$P$_1$ transition in $^{88}$Sr. The reduced reliance on spontaneous emission may allow this adiabatic sweep method to be a useful tool for cooling particles that lack closed cycling transitions, such as molecules.
△ Less
Submitted 6 July, 2017;
originally announced July 2017.
-
Efficient learning with robust gradient descent
Authors:
Matthew J. Holland,
Kazushi Ikeda
Abstract:
Minimizing the empirical risk is a popular training strategy, but for learning tasks where the data may be noisy or heavy-tailed, one may require many observations in order to generalize well. To achieve better performance under less stringent requirements, we introduce a procedure which constructs a robust approximation of the risk gradient for use in an iterative learning routine. Using high-pro…
▽ More
Minimizing the empirical risk is a popular training strategy, but for learning tasks where the data may be noisy or heavy-tailed, one may require many observations in order to generalize well. To achieve better performance under less stringent requirements, we introduce a procedure which constructs a robust approximation of the risk gradient for use in an iterative learning routine. Using high-probability bounds on the excess risk of this algorithm, we show that our update does not deviate far from the ideal gradient-based update. Empirical tests using both controlled simulations and real-world benchmark data show that in diverse settings, the proposed procedure can learn more efficiently, using less resources (iterations and observations) while generalizing better.
△ Less
Submitted 14 October, 2018; v1 submitted 1 June, 2017;
originally announced June 2017.
-
Theory of the crossover from lasing to steady state superradiance
Authors:
D. A. Tieri,
Minghui Xu,
D. Meiser,
J. Cooper,
M. J. Holland
Abstract:
Lasing and steady state superradiance are two phenomena that may appear at first glance to be distinct. In a laser, phase information is maintained by a macroscopic intracavity light field, and the robustness of this phase is what leads to the coherence of the output light. In contrast, the coherence of steady-state superradiant systems derives from the macroscopic collective dipole of a many-atom…
▽ More
Lasing and steady state superradiance are two phenomena that may appear at first glance to be distinct. In a laser, phase information is maintained by a macroscopic intracavity light field, and the robustness of this phase is what leads to the coherence of the output light. In contrast, the coherence of steady-state superradiant systems derives from the macroscopic collective dipole of a many-atom ensemble. In this paper, we develop a quantum theory that connects smoothly between these two extreme limits. We show that lasing and steady-state superradiance should be thought of as the two extreme limits of a continuous crossover. The properties of systems that lie in the superradiance, lasing, and crossover parameter regions are compared. We find that for a given output intensity a narrower linewidth can be obtained by operating closer to the superradiance side of the crossover. We also find that the collective phase is robust against cavity frequency fluctuations in the superradiant regime and against atomic level fluctuations in the lasing regime.
△ Less
Submitted 15 February, 2017;
originally announced February 2017.
-
Semiclassical theory of synchronization-assisted cooling
Authors:
Simon B. Jäger,
Minghui Xu,
Stefan Schütz,
Murray J. Holland,
Giovanna Morigi
Abstract:
We analyse the dynamics leading to radiative cooling of an atomic ensemble confined inside an optical cavity when the atomic dipolar transitions are incoherently pumped and can synchronize. Our study is performed in the semiclassical regime and assumes that cavity decay is the largest rate in the system dynamics. We identify three regimes characterising the cooling. At first hot atoms are individu…
▽ More
We analyse the dynamics leading to radiative cooling of an atomic ensemble confined inside an optical cavity when the atomic dipolar transitions are incoherently pumped and can synchronize. Our study is performed in the semiclassical regime and assumes that cavity decay is the largest rate in the system dynamics. We identify three regimes characterising the cooling. At first hot atoms are individually cooled by the cavity friction forces. After this stage, the atoms' center-of-mass motion is further cooled by the coupling to the internal degrees of freedom while the dipoles synchronize. In the latest stage dipole-dipole correlations are stationary and the center-of-mass motion is determined by the interplay between friction and dispersive forces due to the coupling with the collective dipole. We analyse this asymptotic regime by means of a mean-field model and show that the width of the momentum distribution can be of the order of the photon recoil. Furthermore, the internal excitations oscillate spatially with the cavity standing wave forming an antiferromagnetic-like order.
△ Less
Submitted 6 February, 2017;
originally announced February 2017.
-
Collapse and revival of the monopole mode of a degenerate Bose gas in an isotropic harmonic trap
Authors:
C. J. E. Straatsma,
V. E. Colussi,
M. J. Davis,
D. S. Lobser,
M. J. Holland,
D. Z. Anderson,
H. J. Lewandowski,
E. A. Cornell
Abstract:
We study the monopole (breathing) mode of a finite temperature Bose-Einstein condensate in an isotropic harmonic trap recently developed by Lobser et al. [Nat.~Phys., \textbf{11}, 1009 (2015)]. We observe a nonexponential collapse of the amplitude of the condensate oscillation followed by a partial revival. This behavior is identified as being due to beating between two eigenmodes of the system, c…
▽ More
We study the monopole (breathing) mode of a finite temperature Bose-Einstein condensate in an isotropic harmonic trap recently developed by Lobser et al. [Nat.~Phys., \textbf{11}, 1009 (2015)]. We observe a nonexponential collapse of the amplitude of the condensate oscillation followed by a partial revival. This behavior is identified as being due to beating between two eigenmodes of the system, corresponding to in-phase and out-of-phase oscillations of the condensed and noncondensed fractions of the gas. We perform finite temperature simulations of the system dynamics using the Zaremba-Nikuni-Griffin methodology [J.~Low Temp.~Phys., \textbf{116}, 277 (1999)], and find good agreement with the data, thus confirming the two mode description.
△ Less
Submitted 26 October, 2016; v1 submitted 25 August, 2016;
originally announced August 2016.
-
Supercooling of Atoms in an Optical Resonator
Authors:
Minghui Xu,
Simon B. Jäger,
S. Schütz,
J. Cooper,
Giovanna Morigi,
M. J. Holland
Abstract:
We investigate laser cooling of an ensemble of atoms in an optical cavity. We demonstrate that when atomic dipoles are sychronized in the regime of steady-state superradiance, the motion of the atoms may be subject to a giant frictional force leading to potentially very low temperatures. The ultimate temperature limits are determined by a modified atomic linewidth, which can be orders of magnitude…
▽ More
We investigate laser cooling of an ensemble of atoms in an optical cavity. We demonstrate that when atomic dipoles are sychronized in the regime of steady-state superradiance, the motion of the atoms may be subject to a giant frictional force leading to potentially very low temperatures. The ultimate temperature limits are determined by a modified atomic linewidth, which can be orders of magnitude smaller than the cavity linewidth. The cooling rate is enhanced by the superradiant emission into the cavity mode allowing reasonable cooling rates even for dipolar transitions with ultranarrow linewidth.
△ Less
Submitted 11 December, 2015;
originally announced December 2015.
-
Laser stabilization using saturated absorption in a cavity QED system
Authors:
D. A. Tieri,
J. Cooper,
Bjarke T. R. Christensen,
J. W. Thomsen,
M. J. Holland
Abstract:
We consider the phase stability of a local oscillator (or laser) locked to a cavity QED system comprised of atoms with an ultra-narrow optical transition. The atoms are cooled to millikelvin temperatures and then released into the optical cavity. Although the atomic motion introduces Doppler broadening, the standing wave nature of the cavity causes saturated absorption features to appear, which ar…
▽ More
We consider the phase stability of a local oscillator (or laser) locked to a cavity QED system comprised of atoms with an ultra-narrow optical transition. The atoms are cooled to millikelvin temperatures and then released into the optical cavity. Although the atomic motion introduces Doppler broadening, the standing wave nature of the cavity causes saturated absorption features to appear, which are much narrower than the Doppler width. These features can be used to achieve an extremely high degree of phase stabilization, competitive with the current state-of-the-art. Furthermore, the inhomogeneity introduced by finite atomic velocities can cause optical bistability to disappear, resulting in no regions of dynamic instability and thus enabling a new regime accessible to experiments where optimum stabilization may be achieved.
△ Less
Submitted 26 June, 2015;
originally announced June 2015.