⚛️ quantum physics

Finite Measured Entropy Production Does Not Imply Finite Microscopic Irreversibility

This paper demonstrates that a driven quantum system can exhibit infinite microscopic entropy production while yielding only finite, positive measured entropy production through coarse-grained observations, proving that finite observational data does not necessarily imply finite underlying irreversibility.

Original author: Hieu Nguyen

Published 2026-11-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original author: Hieu Nguyen

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

In the study of how energy moves and changes, scientists have long relied on a simple rule to tell time apart from its reverse: the production of heat and disorder. If you watch a movie of a system running backward, and it looks just as likely to happen as the forward version, the system is reversible. But if the backward version looks impossible—like a shattered cup reassembling itself—then the system has produced entropy, and time has a clear direction. This concept, known as the arrow of time, is usually measured by observing how much energy flows from a hot object to a cold one. In the microscopic world of atoms and light, this flow is tracked by counting individual energy packets as they jump between a system and its surroundings. The more energy that flows, the more certain we are that time is moving forward. However, a lingering question has remained: if our measuring tools are imperfect, can they hide the true scale of this irreversibility? Could a system be generating an infinite amount of disorder while our instruments only see a small, finite amount?

A researcher at Boston University has constructed a theoretical model that answers this question with a definitive yes. They designed a specific setup involving a tiny, driven oscillator—a device that vibrates at a specific frequency—coupled to a thermal environment. In this model, the environment is not a continuous fluid but a stream of individual particles that collide with the oscillator one by one. The researcher simulated a scenario where this environment is cooled down to its absolute lowest possible energy state, known as a vacuum. In this vacuum state, the environment has no energy to give back; it can only receive energy. The oscillator, driven by an external force, continues to pump energy into this vacuum at a steady, finite rate. Because the vacuum cannot return this energy, the microscopic laws of physics dictate that the process is strictly one-way. In the language of probability, the chance of the system running backward is exactly zero. This creates a mathematical singularity where the true, microscopic entropy production becomes infinite.

The core of the study lies in what happens when a realistic detector observes this system. The researcher introduced a detector that mimics the limitations of real-world equipment. This device does not see every single energy packet; it misses some due to inefficiency, it only looks at the system during specific time windows, and it occasionally registers a "click" even when no energy packet is present, a phenomenon known as a dark count. When the researcher analyzed the data recorded by this imperfect counter, they found a startling result. Despite the fact that the system is generating infinite disorder at the microscopic level, the entropy production calculated from the detector's records remains finite and strictly positive. The detector still sees a clear arrow of time; it knows the system is running forward because it sees clicks. However, the magnitude of this arrow is completely disconnected from the true scale of the irreversibility.

The researcher proved that the "missing" entropy, which they call hidden irreversibility, grows without bound as the environment approaches the vacuum state. Specifically, the hidden amount is directly proportional to the rate at which the oscillator dumps energy into the vacuum, divided by the temperature of that environment. As the temperature drops toward zero, this hidden value explodes. The study demonstrates that the only feature of the detector capable of concealing this infinite divergence is the occurrence of false alarms, or dark counts. If the detector never made a mistake and never clicked when nothing happened, the infinite nature of the process would eventually show up in the data, no matter how many real events the detector missed. It is the false positives that blur the line between a system that is merely irreversible and one that is infinitely so. By allowing a click to occur even when the physical process is forbidden in reverse, the false positive masks the fact that the reverse process is impossible, keeping the measured entropy finite.

This finding challenges a common assumption in thermodynamics: that a finite measurement of entropy production guarantees that the underlying physical process is also finite. The paper shows that a small, measurable value does not necessarily mean the system is nearly reversible. It is possible for a system to be generating an infinite amount of entropy while an observer, looking through a flawed lens, sees only a modest, steady signal. The researcher established that the finite signal the observer sees is determined not by the heat flow itself, but by the frequency of the detector's errors. This distinction separates the physical reality of the system from the evidence available to an observer. While the system is undergoing a singular, infinite transformation, the observer's record remains smooth and bounded. The study concludes that without knowing the specific details of the detector's imperfections and the temperature of the environment, one cannot infer the true scale of irreversibility from a measurement record alone. The finite arrow of time seen by the detector is real, but it is a shadow of a much larger, hidden reality.

Technical Summary: Finite Measured Entropy Production Does Not Imply Finite Microscopic Irreversibility

Problem Statement
In stochastic thermodynamics and quantum jump thermodynamics, the production of entropy provides the arrow of time. It is a well-established principle that coarse-grained measurements (imperfect detectors) provide a lower bound on the microscopic entropy production via the data-processing inequality. However, a critical gap remains in understanding the extent to which microscopic irreversibility can be concealed by imperfect observation. Specifically, in quantum systems approaching the vacuum state ( $T_b \to 0$ ), the microscopic entropy production can diverge to infinity due to "absolute irreversibility"—a situation where forward trajectories have positive probability while their time-reversed counterparts have zero probability (e.g., emission is allowed, but absorption from a vacuum is forbidden).

The central question addressed is: Can a physically realizable, imperfect detector register a finite, strictly positive arrow of time while the underlying microscopic entropy production is infinite? Furthermore, is there a quantitative law governing the "hidden" portion of this irreversibility as the bath temperature vanishes?

Methodology
The author constructs a specific driven quantum system and an explicit measurement model to analyze this problem:

Physical System: A single-mode degenerate parametric oscillator driven below threshold. The system is coupled to a thermal reservoir modeled as a sequence of independent bosonic ancillas (collision model). The oscillator is driven by a classical pump at frequency $\omega_p \approx 2\omega_c$ .
Microscopic Dynamics: The system exchanges energy with the bath via emission and absorption jumps. At zero temperature ( $T_b=0$ ), the bath is in a vacuum state, making absorption strictly impossible ( $P_{abs}=0$ ) while emission remains possible ( $P_{emit} > 0$ ). This creates a support mismatch between forward and reverse path measures, leading to infinite microscopic entropy production ( $\Sigma_{full} = \infty$ ).
Stationary State Analysis: The author derives the stationary state of the oscillator, showing it maintains a finite, positive mean photon number and a finite, positive heat current ( $\dot{Q}_0$ ) into the bath even at $T_b=0$ . This current is sustained by the pump creating photon pairs from the vacuum.
Measurement Model (The Counter): A realistic photon counter is defined with three specific imperfections:
- Finite collection efficiency ( $\eta$ ): Not all emitted photons are detected.
- Gated time bins ( $\vartheta$ ): The detector is only "live" with a certain probability.
- Dark counts ( $r$ ): The detector registers false positives with a non-zero probability.
  The detector outputs a binary string (clicks or no clicks) per time bin, representing a coarse-graining of the fine-grained trajectory.

Key Contributions and Results

Divergence of Microscopic Entropy: The paper proves that for the driven parametric oscillator, the fully resolved entropy production $\Sigma_{full}(T, T_b)$ diverges as $T_b \to 0$ . Specifically, $\Sigma_{full} \approx T \dot{Q}_0 / (k_B T_b)$ , where $\dot{Q}_0$ is the stationary vacuum heat current. This divergence arises because the forward process generates events (emissions) that the reverse process (requiring absorption from a vacuum) cannot produce.
Finiteness of Observed Entropy: In contrast to the microscopic divergence, the author proves that the observed relative entropy $\Sigma_Y$ (based on the detector's output) remains finite and strictly positive at vacuum, provided the dark count rate $r > 0$ .
- Theorem 1 establishes rigorous upper and lower bounds for $\Sigma_Y(T, 0)$ .
- Crucially, the paper demonstrates that missed detections (finite $\eta$ ) and gating (finite $\vartheta$ ) cannot conceal the divergence; they merely reduce the signal. Only false positives (dark counts) can conceal the support mismatch. By allowing the reverse law to assign a non-zero probability (via dark counts) to the "click" events observed in the forward law, the singularity is regularized.
The Hidden Irreversibility Law: The paper derives the asymptotic behavior of the hidden entropy production $\Sigma_{hid} = \Sigma_{full} - \Sigma_Y$ as $T_b \to 0$ :
$\Sigma_{hid}(T, T_b) = \frac{T \dot{Q}_0}{k_B T_b} + O(1)$
This result shows that the leading order divergence of the microscopic irreversibility is entirely hidden. The observed entropy production stays bounded, while the hidden part grows inversely with temperature, carrying the full Clausius entropy of the vacuum heat current.
Dependence on Detector Parameters: The observed entropy at vacuum depends on the dark count rate $r$ and collection efficiency $\eta$ . As $r \to 0$ , the observed entropy diverges, confirming that without false positives, the detector would still reveal the infinite irreversibility. However, for any fixed $r > 0$ , the observed entropy is finite.

Significance and Claims
The paper claims to provide the first physically realizable model and explicit detector framework where:

Microscopic entropy production is infinite.
An imperfect detector registers a finite, informative arrow of time.
A quantitative law describes the concealed irreversibility.

The author emphasizes that this result challenges the assumption that a finite measured entropy production implies finite underlying thermodynamic irreversibility. They argue that "finite and informative measurements of entropy production can coexist with infinite microscopic irreversibility."

The distinction between missed events (which preserve the divergence) and false positives (which conceal it) is highlighted as essential. The work suggests that thermodynamic inference from partial observations has fundamental limits: a small measured value cannot establish near-reversibility, and a finite measurement cannot guarantee finite underlying entropy production without additional knowledge of the detector's response and the bath's temperature. The paper concludes by noting potential applications in stochastic inflation and quantum information, where coarse-graining might similarly conceal divergent irreversibility.

Technical Summary: Finite Measured Entropy Production Does Not Imply Finite Microscopic Irreversibility

More like this