arXiv subjects

David Roberts

Publications and source records attributed to David Roberts.

At least 19 recordsLinked to original sources

arXiv2026

Bosonic codes from compact phase spaces

We present the algebraic structure of bosonic quantum error-correcting codes on genus-two Riemann surfaces. We explicitly construct the code words as automorphic forms and analytically generate the full tower of code spaces at all weights. We prove a fundamental no-go theorem: for any genus greater than one, the stabilizer group is non-amenable, forcing a strictly positive spectral gap in the stabilizer Hamiltonian. Consequently, no normalizable quantum state can satisfy all stabilizer conditions. This sharply contrasts with standard Gottesman-Kitaev-Preskill (GKP) codes, where the amenability of the stabilizer group $\mathbb{Z}^2$ permits approximate code words with arbitrary precision.

quant-ph↗

arXiv2026

Thermalizing Stochastic Programs

We present a set of tools for mapping general stochastic programs to thermodynamic hardware designed for energy-efficient stochastic sampling. Given a target stochastic program expressed as a Directed Factor Graph (DFG) of stochastic channels, or equivalently as a Parametrized Stochastic Circuit (PSC), we first introduce a method to approximately compile each factor in the DFG to an Energy-Based Model (EBM) that is native to the hardware. We then analyze how the error of the compiled DFG accumulates from the per-factor errors, and introduce two training refinements, context matching and trajectory-level REINFORCE post-training, which can reduce the residual error left by training each factor in isolation. The \texttt{thermalizers} framework takes a stochastic program expressed in the \texttt{torx} library and replaces its factors with thermodynamic kernels implemented and sampled using the \texttt{thrml} library. We demonstrate it on several example applications, including a market simulator that learns the joint day-to-day dynamics of a panel of financial time series from recorded market history alone, a probabilistic model from mathematical ecology, Gibbs sampling of an EBM the hardware cannot natively express, and a sequential Bayesian design loop over a Gaussian stochastic circuit.

cs.ET↗

arXiv2026

When Your LLM Reaches End-of-Life: A Framework for Confident Model Migration in Production Systems

We present a framework for migrating production Large Language Model (LLM) based systems when the underlying model reaches end-of-life or requires replacement. The key contribution is a Bayesian statistical approach that calibrates automated evaluation metrics against human judgments, enabling confident model comparison even with limited manual evaluation data. We demonstrate this framework on a commercial question-answering system serving 5.3M monthly interactions across six global regions; evaluating correctness, refusal behavior, and stylistic adherence to successfully identify suitable replacement models. The framework is broadly applicable to any enterprise deploying LLM-based products, providing a principled, reproducible methodology for model migration that balances quality assurance with evaluation efficiency. This is a capability increasingly essential as the LLM ecosystem continues to evolve rapidly and organizations manage portfolios of AI-powered services across multiple models, regions, and use cases.

cs.AI↗

arXiv2025

Improved diffusive approximation of Markov jump processes close to equilibrium

Diffusive approximations of Markov jump processes often fail to accurately capture large fluctuations. This is confounding, as the rare events triggered by these large fluctuations, such as the failure of electronic memories, are often the object of interest. In this paper we present an improved diffusive approximation, extending a method previously limited to equilibrium systems. Using new tools from stochastic thermodynamics, we prove its validity to linear order in departures from equilibrium and demonstrate its superior accuracy over the Kramers-Moyal expansion in predicting both steady-state and transient properties, including the error rate of a non-equilibrium electronic memory.

cond-mat.mes-hall↗

arXiv2025

A Limits Study of Memory-side Tiering Telemetry

Increasing workload demands and emerging technologies necessitate the use of various memory and storage tiers in computing systems. This paper presents results from a CXL-based Experimental Memory Request Logger that reveals precise memory access patterns at runtime without interfering with the running workloads. We use it for software emulation of future memory telemetry hardware. By combining reactive placement based on data address monitoring, proactive data movement, and compiler hints, a Hotness Monitoring Unit (HMU) within memory modules can greatly improve memory tiering solutions. Analysis of page placement using profiled access counts on a Deep Learning Recommendation Model (DLRM) indicates a potential 1.94x speedup over Linux NUMA balancing tiering, and only a 3% slowdown compared to Host-DRAM allocation while offloading over 90% of pages to CXL memory. The study underscores the limitations of existing tiering strategies in terms of coverage and accuracy, and makes a strong case for programmable, device-level telemetry as a scalable and efficient solution for future memory systems.

cs.OS↗

arXiv2024

Hidden time-reversal in driven XXZ spin chains: exact solutions and new dissipative phase transitions

We show that several models of interacting XXZ spin chains subject to boundary driving and dissipation possess a subtle kind of time-reversal symmetry, making their steady states exactly solvable. We focus on a model with a coherent boundary drive, showing that it exhibits a unique continuous dissipative phase transition as a function of the boundary drive amplitude. This transition has no analogue in the bulk closed system, or in incoherently driven models. We also show the steady state magnetization exhibits a surprising fractal dependence on interaction strength, something previously associated with less easily measured infinite-temperature transport quantities (the Drude weight). Our exact solution also directly yields driven-dissipative double-chain models that have pure, entangled steady states that are also current carrying.

quant-ph↗

arXiv2023

Exact solution of the infinite-range dissipative transverse-field Ising model

The dissipative variant of the Ising model in a transverse field is one of the most important models in the analysis of open quantum many-body systems, due to its paradigmatic character for understanding driven-dissipative quantum phase transitions, as well as its relevance in modelling diverse experimental platforms in atomic physics and quantum simulation. Here, we present an exact solution for the steady state of the transverse-field Ising model in the limit of infinite-range interactions, with local dissipation and inhomogeneous transverse fields. Our solution holds despite the lack of any collective spin symmetry or even permutation symmetry. It allows us to investigate first- and second-order dissipative phase transitions, driven-dissipative criticality, and captures the emergence of a surprising "spin blockade" phenomenon. The ability of the solution to describe spatially-varying local fields provides a new tool to study disordered open quantum systems in regimes that would be extremely difficult to treat with numerical methods.

quant-ph↗

arXiv2022

Contact-Free Simultaneous Sensing of Human Heart Rate and Canine Breathing Rate for Animal Assisted Interactions

Animal Assisted Interventions (AAIs) involve pleasant interactions between humans and animals and can potentially benefit both types of participants. Research in this field may help to uncover universal insights about cross-species bonding, dynamic affect detection, and the influence of environmental factors on dyadic interactions. However, experiments evaluating these outcomes are limited to methodologies that are qualitative, subjective, and cumbersome due to the ergonomic challenges related to attaching sensors to the body. Current approaches in AAIs also face challenges when translating beyond controlled clinical environments or research contexts. These also often neglect the measurements from the animal throughout the interaction. Here, we present our preliminary effort toward a contact-free approach to facilitate AAI assessment via the physiological sensing of humans and canines using consumer-grade cameras. This initial effort focuses on verifying the technological feasibility of remotely sensing the heart rate signal of the human subject and the breathing rate signal of the dog subject while they are interacting. Small amounts of motion such as patting and involuntary body shaking or movement can be tolerated with our custom designed vision-based algorithms. The experimental results show that the physiological measurements obtained by our algorithms were consistent with those provided by the standard reference devices. With further validation and expansion to other physiological parameters, the presented approach offers great promise for many scenarios from the AAI research space to veterinary, surgical, and clinical applications.

eess.SP↗

arXiv2022

Competition between two-photon driving, dissipation and interactions in bosonic lattice models: an exact solution

We present an exact solution in arbitrary dimensions for the steady states of a class of quantum driven-dissipative bosonic models, where a set of modes is subject to arbitrary two-photon driving, single-photon loss and a global Hubbard (or Kerr)-like interaction. Our solutions reveal a wealth of striking phenomena, including the emergence of dissipative phase transitions, nontrivial mode competition physics and symmetry breaking, and the stabilization of many-body $SU(1,1)$ pair coherent states. Our exact solutions enable the description of spatial correlations, and are fully valid in regimes where traditional mean-field and semiclassical approaches break down.

quant-ph↗

arXiv2021

Unconditional Fock state generation using arbitrarily weak photonic nonlinearities

We present a new mechanism that harnesses extremely weak Kerr-type nonlinearities in a single driven cavity to deterministically generate single photon Fock states, and more general photon-blockaded states. Our method is effective even for nonlinearities that are orders-of-magnitude smaller than photonic loss. It is also completely distinct from so-called unconventional photon blockade mechanisms, as the generated states are non-Gaussian, exhibit a sharp cut-off in their photon number distribution, and can be arbitrary close to a single-photon Fock state. Our ideas require only standard linear and parametric drives, and is hence compatible with a variety of different photonic platforms.

quant-ph↗

arXiv2020

Hidden time-reversal symmetry, quantum detailed balance and exact solutions of driven-dissipative quantum systems

Driven-dissipative quantum systems generically do not satisfy simple notions of detailed balance based on the time symmetry of correlation functions. We show that such systems can nonetheless exhibit a hidden time-reversal symmetry which most directly manifests itself in a doubled version of the original system prepared in an appropriate entangled thermofield double state. This hidden time-reversal symmetry has a direct operational utility: it provides a general method for finding exact solutions of non-trivial steady states. Special cases of this approach include the coherent quantum absorber and complex-$P$ function methods from quantum optics. We also show that hidden TRS has observable consequences even in single-system experiments, and can be broken by the non-trivial combination of nonlinearity, thermal fluctuations, and driving. To illustrate our ideas, we analyze concrete examples of driven qubits and nonlinear cavities. These systems exhibit hidden time-reversal symmetry but not conventional detailed balance.

quant-ph↗

arXiv2019

Driven-dissipative quantum Kerr resonators: new exact solutions, photon blockade and quantum bistability

We present a new approach for deriving exact, closed-form solutions for the steady state of a wide class of driven-dissipative nonlinear resonators that is distinct from more common positive-$P$ function methods. Our method generalizes the coherent quantum absorber approach of Stannigel et al. to include nonlinear driving and dissipation, and relies crucially on exploiting the Segal-Bargmann representation of Fock space. Our solutions and method reveal a wealth of previously unexplored observable phenomena in these systems, including new generalized photon-blockade and anti-blockade effects, and an infinite number of new parameter choices that yield quantum-bistability.

quant-ph↗

arXiv2019

Noise amplification at spin-glass bottlenecks of quantum annealing: a solvable model

To gain better insight into the complexity theory of quantum annealing, we propose and solve a class of spin systems which contain bottlenecks of the kind expected to dominate the runtime of quantum annealing as it tries to solve difficult optimization problems. We uncover a noise amplification effect at these bottlenecks, whereby tunneling rates caused by flux-qubit noise scale in proportion to the number of qubits $N$ in the limit that $N\to \infty$. By solving the incoherent annealing dynamics exactly, we find a wide range of regimes where the probability that a quantum annealer remains in the ground-state upon exiting the bottleneck is close to one-half. We corroborate our analysis with detailed simulations of the performance of the D-Wave 2X quantum annealer on our class of computational problems.

quant-ph↗

arXiv2018

Hypergeometric supercongruences

We discuss two related principles for hypergeometric supercongrences, one related to accelerated convergence and the other to the vanishing of Hodge numbers. This is an extended abstract of a talk given at the workshop "Hypergeometric motives and Calabi-Yau differential equations" at the Mathematical Research Institute (MATRIX) of The University of Melbourne in Creswick, Australia in January of 2017.

math.NT↗

arXiv2017

Interactive Learning from Policy-Dependent Human Feedback

This paper investigates the problem of interactively learning behaviors communicated by a human teacher using positive and negative feedback. Much previous work on this problem has made the assumption that people provide feedback for decisions that is dependent on the behavior they are teaching and is independent from the learner's current policy. We present empirical results that show this assumption to be false -- whether human trainers give a positive or negative feedback for a decision is influenced by the learner's current policy. Based on this insight, we introduce {\em Convergent Actor-Critic by Humans} (COACH), an algorithm for learning from policy-dependent feedback that converges to a local optimum. Finally, we demonstrate that COACH can successfully learn multiple behaviors on a physical robot.

cs.AI↗

arXiv2016

Monopole-Enriched Groundstates of Two-Dimensional BCS Models

We extract the macroscopic characteristics of the groundstate sectors of two dimensional Bose-condensed BCS models with a fixed number of magnetic flux quanta, on length scales much larger than the characteristic size of the Cooper-pair bound states. We show that this reduces to the problem of computing the moduli space of gauge-inequivalent solutions to the Gross-Pitaevskii equations on a nontrivial fibre bundle. Inspired in part by the physical arguments of Oshikawa and Senthil in Fractionalization, Topological Order, and Quasiparticle Statistics, we extract a large class of groundstates from this moduli space.

cond-mat.supr-con↗

arXiv2016

The Total Green's Function of a Non-Interacting System

Despite its centrality in the mathematical structure of perturbative many-body theory, the total Green's function for the many-body time-dependent Schrodinger equation has been ignored for decades, superseded by single-particle Green's functions, for which a vast portion of the literature has been devoted. In this paper, we give the first computation the total Green's function for the time-dependent Schrodinger equation for a non-interacting system of identical particles, setting the stage for a fresh interpretation of perturbative many-body physics.

cond-mat.str-el↗

arXiv2008

The inner automorphism 3-group of a strict 2-group

Any group $G$ gives rise to a 2-group of inner automorphisms, $\mathrm{INN}(G)$. It is an old result by Segal that the nerve of this is the universal $G$-bundle. We discuss that, similarly, for every 2-group $G_{(2)}$ there is a 3-group $\mathrm{INN}(G_{(2)})$ and a slightly smaller 3-group $\mathrm{INN}_0(G_{(2)})$ of inner automorphisms. We describe these for $G_{(2)}$ any strict 2-group, discuss how $\mathrm{INN}_0(G_{(2)})$ can be understood as arising from the mapping cone of the identity on $G_{(2)}$ and show that its underlying 2-groupoid structure fits into a short exact sequence $G_{(2)} \to \mathrm{INN}_0(G_{(2)}) \to ΣG_{(2)}$. As a consequence, $\mathrm{INN}_0(G_{(2)})$ encodes the properties of the universal $G_{(2)}$ 2-bundle.

math.CT↗