Quantum thermodynamics with uncertain equilibrium
Munan Zhang, Kun Fang
Abstract
The resource-theoretic approach to quantum thermodynamics assumes complete knowledge of the thermal equilibrium against which thermodynamic resources are defined. In practice, however, this state is determined by the system Hamiltonian and the bath temperature, neither of which is known with perfect precision. We develop a framework in which the equilibrium reference is specified by a set of candidate states reflecting this uncertainty. Under a generic geometric condition, we prove a no-go theorem that sharply limits athermality ``purification'': conversion from an uncertain athermality resource to a definite target is either trivial or impossible, with no room for tradeoff. We then introduce two complementary battery models: a clean battery with a precisely known equilibrium state and a dirty battery with an uncertain one. For both models, we derive exact one-shot entropic characterizations of work extraction and work of formation in terms of standard min- and max-relative entropies and new subspace-constrained variants. In the asymptotic regime, both models exhibit a strong form of thermodynamic irreversibility. In particular, we give a simple and explicit example in which, in the clean-battery model, work is required to form a state but no work can be extracted from it, in direct analogy with bound entanglement, whereas in the dirty-battery model, work can be extracted but formation requires infinite work cost. These phenomena persist even under arbitrarily small uncertainty, showing that equilibrium uncertainty is not a minor perturbation of the standard theory but a qualitatively new ingredient that reshapes the fundamental limits of thermodynamic resource interconversion.
AI Impact Assessments
(3 models)Scientific Impact Assessment: "Quantum thermodynamics with uncertain equilibrium"
1. Core Contribution
This paper introduces a framework for quantum thermodynamics where the equilibrium (Gibbs) state is not precisely known — reflecting practical uncertainty in system Hamiltonians and bath temperatures. The main contributions are:
2. Methodological Rigor
The paper is technically rigorous. The proofs are detailed and complete (Appendix A spans 12 pages of careful derivations). Key strengths include:
One minor concern: the framework adopts a worst-case (minimax) notion of success, which is natural but potentially conservative. An average-case or Bayesian formulation might yield different conclusions, though this is acknowledged implicitly.
3. Potential Impact
Within quantum thermodynamics: This work challenges a foundational assumption that has been largely unquestioned. The demonstration that equilibrium uncertainty is qualitatively (not just quantitatively) different from the standard theory is a significant conceptual advance. The analogy with bound entanglement is particularly striking and could stimulate new research directions.
Quantum information theory: The new subspace-constrained relative entropies (D^E_min,ε and D^E_max,ε) are mathematically interesting objects. Their operational interpretation as constrained hypothesis testing problems (where the test must be "calibration-free") could find applications beyond thermodynamics. The connection to cone-restricted information theory (George-Chitambar 2024) and generalized AEP (Fang-Fawzi-Fawzi 2024) enriches the entropic landscape.
Experimental relevance: The motivation is grounded in practical limitations — Hamiltonian learning, temperature estimation, and parameter drift are real experimental concerns. The results suggest that resource-theoretic predictions may be overly optimistic unless equilibrium uncertainty is accounted for.
Broader resource theories: The paper explicitly notes open questions about whether similar phenomena arise in resource theories of coherence, asymmetry, and reference frames. This could catalyze a systematic re-examination of resource theories with uncertain free states.
4. Timeliness & Relevance
The paper addresses a genuine gap. The resource-theoretic approach to thermodynamics has matured significantly over the past decade, yet the assumption of perfect equilibrium knowledge has remained unquestioned. Recent work on black-box thermodynamics (Watanabe-Takagi 2024, 2026; Šafránek et al. 2023) has begun relaxing assumptions about the nonequilibrium state, but this paper is the first to systematically address uncertainty in the *equilibrium* reference itself. Given growing interest in quantum error correction, calibration protocols, and Hamiltonian learning in NISQ and early fault-tolerant devices, this contribution is well-timed.
5. Strengths & Limitations
Key Strengths:
Notable Limitations:
6. Additional Observations
The conceptual parallel with bound entanglement is intellectually stimulating but deserves careful qualification: in entanglement theory, bound entanglement is a property of specific states, while here the "bound athermality" arises from *uncertainty* about which state one has. This distinction is important and could be explored further.
The paper's results on the resource theory of asymmetric distinguishability (mentioned in the Discussion) represent an additional avenue of impact, as that resource theory has been gaining traction as a unifying framework.
Generated Apr 16, 2026
Comparison History (48)
Paper 2 introduces a fundamentally new theoretical framework for quantum thermodynamics under equilibrium uncertainty, proving a sharp no-go theorem and revealing qualitatively new phenomena (analogues of bound entanglement in thermodynamics, persistent irreversibility under arbitrarily small uncertainty). This challenges and extends the foundational resource-theoretic framework used broadly across quantum information and thermodynamics. Paper 1 proposes a useful engineering protocol for multinode entanglement but is more incremental. Paper 2's broader conceptual impact across multiple fields—quantum information, thermodynamics, and resource theories—gives it higher potential scientific impact.
Paper 1 presents a novel experimental observation of complex nonlinear dynamics at very low (microwatt) power levels. Its direct applications in high-precision sensing, microwave signal generation, and neuromorphic computing provide immediate, tangible impacts across applied physics and engineering, arguably offering broader and faster real-world translation than the purely theoretical framework of Paper 2.
Paper 2 likely has higher near-term scientific impact: it reports experimental observation of attractor transitions, chaos, and large amplified spectral response in an active magnon-polariton platform at microwatt powers, enabling concrete applications (low-power nonlinear microwave sources, sensing, neuromorphic computing). The combination of calibrated stability analysis with experimentally mapped phase behavior suggests solid rigor and broad relevance across nonlinear dynamics, spintronics, microwave photonics, and applied physics. Paper 1 is conceptually novel and rigorous in quantum resource theory, but is more specialized and further from immediate experimental validation.
Paper 1 introduces a fundamentally new theoretical framework for quantum thermodynamics under equilibrium uncertainty, proving no-go theorems and discovering novel phenomena (analogues of bound entanglement, irreversibility under arbitrarily small uncertainty) that reshape foundational understanding. Its conceptual contributions—showing uncertainty is qualitatively distinct, not a perturbation—have broad implications across quantum information and thermodynamics. Paper 2, while technically impressive in demonstrating exascale simulation for QPU benchmarking, is more incremental: it validates a specific processor at specific scales, with results that will be superseded as hardware improves. Paper 1's theoretical insights have longer-lasting, broader impact.
Paper 2 likely has higher near-term scientific impact: it delivers concrete, timely benchmarking methodology for a 98-qubit QPU using exascale simulation, producing quantitative, actionable thresholds (noise-tolerant vs random regimes) relevant to industry and experimental roadmaps. Its real-world applicability is immediate (QPU validation, performance tracking, HPC–QC co-design) and impacts multiple communities (quantum hardware, algorithms, HPC). Paper 1 is highly novel and rigorous theoretically, but its impact may be narrower and longer-term, with fewer direct experimental touchpoints.
Paper 2 addresses a highly timely and critical bottleneck in quantum computing: quantum state learning and characterization. By providing practical, logarithmic-depth algorithms for the average case and fundamental exponential lower bounds for the worst case, it offers immediate utility for near-term quantum device verification (QCVV) alongside deep theoretical insights. Paper 1 is an elegant theoretical advance in quantum thermodynamics, but Paper 2 has broader and more immediate applicability across the rapidly growing quantum information and computing ecosystem.
Paper 1 introduces a fundamentally new framework for quantum thermodynamics under equilibrium uncertainty, proving no-go theorems and discovering novel phenomena (analogues of bound entanglement, irreversibility under arbitrarily small uncertainty) that reshape foundational understanding. Its results are qualitatively surprising—showing uncertainty isn't a perturbation but a structural change—with broad implications across quantum information and thermodynamics. Paper 2 makes a solid contribution unifying cooperative emission scaling laws, but is more incremental, clarifying known phenomena rather than revealing fundamentally new physics. Paper 1's conceptual depth and cross-field relevance give it higher impact potential.
Paper 2 likely has higher impact: it introduces a fundamentally new ingredient (equilibrium uncertainty) into quantum thermodynamic resource theories, proves a sharp no-go theorem, and provides exact one-shot and asymptotic characterizations with striking irreversibility/“bound-thermo” phenomena persisting under arbitrarily small uncertainty. This is conceptually deep, broadly relevant across quantum information/thermo/resource theories, and methodologically rigorous. Paper 1 is timely and practically motivated (LLM-guided VQA policy search), but impact may be narrower/less durable given rapid iteration in AutoML/LLM tooling and dependence on near-term quantum optimization benchmarks.
Paper 1 introduces a foundational framework that reshapes quantum thermodynamics under realistic uncertainty, offering rigorous no-go theorems and discovering novel irreversibility phenomena. In contrast, Paper 2 presents a useful but largely incremental heuristic algorithm for quantum optimal control with competitive, rather than groundbreaking, empirical results. Paper 1's profound theoretical implications offer deeper novelty and a more lasting fundamental scientific impact.
Paper 2 introduces a fundamentally new theoretical framework for quantum thermodynamics under equilibrium uncertainty, proving a no-go theorem and revealing qualitatively new phenomena (analogies to bound entanglement, irreversibility under arbitrarily small uncertainty). This reshapes foundational understanding of thermodynamic resource theories with broad implications. Paper 1, while technically impressive in developing quantum decoders for optimization, explicitly acknowledges it falls short of quantum advantage due to a classical enhancement achieving a precise tie, limiting its immediate impact.
Paper 1 offers a fundamental theoretical advance: it relaxes a core idealization (perfect equilibrium knowledge) in quantum thermodynamic resource theories, proves a sharp no-go result, and derives exact one-shot and asymptotic characterizations with qualitatively new irreversibility/bound-resource phenomena persisting under arbitrarily small uncertainty. This is high-novelty, broadly relevant across quantum information/thermo foundations, and likely to reframe limits used by many subsequent works. Paper 2 is timely and application-oriented, but impact may be narrower and more incremental given dependence on curated design spaces, benchmarks, and rapidly evolving LLM tooling.
Paper 1 addresses foundational questions in quantum thermodynamics by introducing realistic uncertainty into equilibrium states, deriving a novel no-go theorem, and demonstrating phenomena analogous to bound entanglement. This establishes qualitatively new fundamental limits in thermodynamic resource interconversion. While Paper 2 offers a useful heuristic algorithm for quantum optimal control, Paper 1's theoretical breakthroughs have a deeper and more lasting conceptual impact on the foundational understanding of quantum physics.
Paper 2 introduces a fundamentally new framework for quantum thermodynamics under equilibrium uncertainty, proving a sharp no-go theorem and revealing qualitatively new phenomena (analogies to bound entanglement, persistent irreversibility under arbitrarily small uncertainty). This reshapes foundational understanding of thermodynamic resource theories with broad implications. Paper 1, while technically strong in developing quantum decoders for optimization, explicitly acknowledges it falls short of achieving quantum advantage, with classical algorithms matching its performance. Paper 2's results are more definitive and foundational, with broader cross-field impact.
Paper 1 addresses a critical, immediate bottleneck in the development of fault-tolerant quantum computers: scalable and fast quantum error correction. By proposing a high-performance, hardware-friendly local cellular automaton decoder (SCALA), it offers substantial practical applications and hardware implementation potential. Paper 2, while offering profound theoretical insights into quantum thermodynamics and resource interconversion, is highly foundational. The timeliness, real-world applicability, and engineering relevance of Paper 1 give it a higher potential for broad scientific and technological impact in the rapidly advancing field of quantum computing.
Paper 2 is more conceptually novel and broadly impactful: it relaxes a core assumption of resource-theoretic quantum thermodynamics (known equilibrium), proves a general no-go theorem, and provides exact one-shot and asymptotic characterizations with new entropy variants and battery models. Its results reshape fundamental limits and connect to bound-entanglement-like irreversibility, relevant across thermodynamics, information theory, and foundations, with timeliness for realistic imperfect-knowledge settings. Paper 1 is practically useful for experiments on monitored circuits, but is more application/engineering-oriented and narrower in scope, and ML-based phase classification may face generalization/interpretability concerns.
Paper 2 demonstrates the first experimental realization of 2D continuous-variable cluster states in the microwave domain using 191 frequency modes, which is a significant milestone for measurement-based quantum computing with superconducting circuits. This bridges CV quantum optics and microwave quantum technologies, with direct implications for scalable quantum computing architectures. While Paper 1 presents elegant theoretical results on quantum thermodynamics under equilibrium uncertainty with interesting analogies to bound entanglement, Paper 2's experimental breakthrough has broader immediate impact across quantum computing, quantum networks, and superconducting circuit communities, and opens concrete pathways toward practical quantum information processing.
Paper 1 provides rigorous theoretical foundations for belief propagation in tensor networks—a widely used computational tool in quantum many-body physics—establishing sharp criteria for when BP succeeds or fails, with direct connections to physical correlation functions. This impacts a broad community using tensor network methods and belief propagation across condensed matter physics, quantum information, and computational physics. Paper 2 makes elegant contributions to quantum thermodynamics resource theory under uncertainty, but addresses a more niche setting. Paper 1's combination of rigorous theory, practical computational relevance, and numerical validation gives it broader impact potential.
Paper 2 is more novel and broadly impactful: it relaxes a core assumption in resource-theoretic quantum thermodynamics (known equilibrium) and shows qualitatively new effects (sharp no-go theorem, strong irreversibility, bound-entanglement-like behavior) that persist under arbitrarily small uncertainty. It provides rigorous one-shot and asymptotic characterizations with new entropic quantities and clear implications for realistic thermodynamic modeling where Hamiltonians/temperatures are imperfectly known. Paper 1 is solid and useful for optimizing entanglement harvesting, but is narrower in scope and more incremental within an established framework.
Paper 1 is more novel and broadly impactful: it extends quantum thermodynamic resource theory to realistic equilibrium uncertainty, proves a general no-go theorem, introduces new battery models, and provides exact one-shot entropic characterizations plus striking irreversibility/bound-resource analogs that persist under arbitrarily small uncertainty. This reshapes foundational limits and could influence multiple areas (resource theories, quantum info, statistical mechanics, metrology). Paper 2 is timely and practically relevant for quantum memories, but is narrower in scope and appears more engineering/optimization-focused with impact mainly in a specific platform.
Paper 1 is more novel and broadly impactful: it extends quantum thermodynamic resource theories to realistic equilibrium uncertainty, proves a sharp no-go theorem, and develops two battery models with exact one-shot and asymptotic characterizations, revealing new irreversibility phenomena (bound-thermo analogs) persisting under small uncertainty. This is timely and potentially foundational for nanoscale thermodynamics and experimental implementations with imperfect calibration. Paper 2 mainly analyzes coherence measures (Tsallis α) within a well-studied algorithm (Grover), yielding complementarity relations but with narrower conceptual and application scope and likely incremental impact.