Macroscopicity and observational deficit in states, operations, and correlations

This paper establishes a unified inferential framework for macroscopicity based on observational deficit and Bayesian retrodiction, which defines macroscopic entropy, formulates a resource theory of microscopicity that generalizes existing theories, and characterizes quantum correlations under observational constraints.

The Gist

Imagine you are looking at a high-resolution photograph of a bustling city. You can see every face, every license plate, and every crack in the pavement. This is the microscopic view—the complete, detailed truth of the system.

Now, imagine you put on a pair of very foggy glasses. Suddenly, you can't see individual people anymore. You only see blurry blobs of color moving in different directions. You can tell where the traffic is heavy, but you can't tell who is driving which car. This is the macroscopic view.

This paper is about understanding the gap between those two views. It asks: What do we lose when we look at the world through "foggy glasses"? And can we ever get the original picture back?

Here is a breakdown of the paper's big ideas using simple analogies.

1. The Foggy Glasses (Coarse-Graining)

In the quantum world, scientists usually deal with perfect, detailed information. But in real life, our instruments (and our eyes) are limited. We can't measure everything at once.

The authors call this limitation "Coarse-Graining." Think of it like taking a high-definition video and compressing it into a low-resolution GIF. You lose the fine details, but you keep the general motion.

The paper introduces a new way to think about this: Observational Deficit.

• The Analogy: Imagine you take a photo of a messy room (the microscopic state). Then, you take a blurry photo of it (the macroscopic measurement).
• The Deficit: The "Observational Deficit" is a score that tells you how much information was lost in the blur. If the score is zero, the blur didn't hide anything important. If the score is high, the blur hid a lot of secrets.

2. The "Perfect Guess" (Bayesian Retrodiction)

Usually, we think of science as moving forward: Cause $\rightarrow$ Effect. But this paper looks backward: Effect $\rightarrow$ Cause. This is called Retrodiction.

Imagine you walk into a kitchen and see a broken egg on the floor.

• Microscopic view: You know exactly how the egg fell, the speed, the angle, and the texture of the shell.
• Macroscopic view: You just see "a broken egg."

The paper asks: Based on the "broken egg" you see, what is the most logical guess about how it got there?

They use a mathematical tool called the Petz Recovery Map (named after a famous physicist). Think of this as a "Smart AI" that tries to reconstruct the original state from the blurry data.

• If the AI can perfectly reconstruct the original state, the state is "Macroscopic." It means the state was simple enough that the foggy glasses didn't hide anything.
• If the AI fails, the state is "Microscopic" (in the sense that it holds hidden details the observer can't see).

3. The "Inferential Reference Frame" (The Observer's Map)

One of the coolest ideas in the paper is the Inferential Reference Frame.

Think of an observer not just as a person, but as a specific map they are using to navigate the world.

• The Map: This map is defined by two things:
  1. The Prior: What the observer expects to see (their background knowledge).
  2. The Measurement: What tools they have to look at the world (their foggy glasses).

The paper proves that for any specific map (any specific observer with specific tools), there is a unique "Best Map" (called the Maximal Projective Post-Processing).

• Analogy: Imagine you are trying to describe a forest. You have a blurry camera. The "Best Map" is the most detailed description of the forest that your blurry camera could possibly capture. Anything more detailed than that is invisible to you.

This "Best Map" acts like a reference frame. It defines what is "real" for that specific observer.

4. The Resource Theory of "Microscopicity"

In physics, a "Resource Theory" is a way to say: "Some things are valuable, and some things are free."

• Free Stuff: Things that are easy to make or don't require special effort.
• Valuable Stuff: Things that are hard to make and require special resources.

The authors flip the script. They say:

• Macroscopic States are "Free": These are the simple, blurry states that anyone can see. They are the "default" state of the world.
• Microscopic States are "Resources": These are the complex, hidden states that require special, high-resolution tools to detect.

They create a hierarchy of "operations" (actions you can take):

1. Macroscopic Operations: Actions that keep things blurry. You can't turn a blurry photo into a sharp one just by shaking the camera.
2. Microscopic Operations: Actions that can reveal hidden details.

This unifies many other famous theories in physics (like Coherence and Asymmetry) under one big umbrella. It's like realizing that "heat," "electricity," and "sound" are all just different forms of energy.

5. Quantum Correlations (The "Spooky" Connection)

Finally, the paper looks at Quantum Entanglement (where two particles are linked across the universe). Usually, we think entanglement is an absolute fact: "These two particles are connected."

But this paper argues: Entanglement depends on who is looking.

• The Analogy: Imagine two people, Alice and Bob, are whispering secrets to each other.
  • If you have a super-sensitive microphone (a perfect observer), you hear the secrets. You see the connection.
  • If you have a cheap, static-filled radio (a macroscopic observer), you hear nothing but noise. To you, Alice and Bob seem unconnected.

The authors introduce "Observational Discord." This measures how much "spooky connection" is visible to a specific observer with specific limitations.

• If your "foggy glasses" are too thick, the quantum connection disappears from your view.
• This means quantum correlations aren't just "there" or "not there." They are context-dependent. They exist relative to the observer's ability to see them.

Summary: Why Does This Matter?

This paper gives us a new way to understand the universe:

1. Irreversibility: It explains why time seems to move forward (entropy increases). It's because our "foggy glasses" (measurements) lose information. We can't un-blur the photo.
2. The Observer Matters: Reality isn't just a fixed stage; it's a stage viewed through a specific lens. What looks "quantum" to one person might look "classical" to another.
3. Unified Theory: It connects many different branches of quantum physics (thermodynamics, information, symmetry) into one single language of "what can be seen and what is hidden."

In short, the paper tells us that macroscopic reality is the result of what we can successfully infer from our limited measurements. The "blur" isn't a mistake; it's the fundamental nature of how we experience the quantum world.

Technical Summary

1. Problem Statement

The paper addresses the fundamental challenge of explaining the emergence of macroscopic irreversibility (e.g., the Second Law of Thermodynamics) from microscopic reversible quantum dynamics.

• The Gap: Standard von Neumann entropy remains invariant under unitary evolution, failing to capture the entropy increase observed in macroscopic systems.
• The Limitation of Existing Models: While "observational entropy" has been proposed to bridge this gap, previous formulations were often limited to specific cases (e.g., Projection-Valued Measures or uniform priors). There was a lack of a unified, inferential framework that could handle arbitrary Positive Operator-Valued Measures (POVMs) and arbitrary quantum priors (representing the observer's prior knowledge).
• Core Question: How can we rigorously define "macroscopic states," quantify the loss of microscopic information (irretrodictability), and understand how quantum correlations (like entanglement and discord) depend on the observer's measurement capabilities?

2. Methodology

The authors develop a unified framework based on Quantum Statistical Sufficiency and Bayesian Retrodiction. The methodology proceeds through the following logical steps:

1. Inferential Framework: They define macroscopicity not as an intrinsic property of a state, but relative to an Observational Reference Frame, defined by a pair $(P, \gamma)$, where $P$ is a POVM (measurement) and $\gamma$ is a prior state.
2. Coarse-Graining Maps: They construct a "coarse-graining map" $C_{P,\gamma}$ as the composition of a measurement channel $M_P$ and its associated Petz recovery map $R_{M_P, \gamma}$. This map represents the optimal Bayesian inference (retrodiction) an observer can make given measurement outcomes and a prior.
3. Observational Deficit: They introduce the observational deficit $\delta_P(\rho \| \gamma)$, defined as the difference between the relative entropy of the state and the relative entropy of the coarse-grained state. This quantifies the "irretrodictability" or information loss.
4. Resource Theory: They formalize a Resource Theory of Microscopicity, treating macroscopic states (zero deficit) as "free states" and defining a hierarchy of "free operations" that cannot generate microscopic details.
5. Correlation Analysis: They apply this framework to bipartite systems to define Observational Discord, analyzing how correlations vanish or persist under specific observational constraints.

3. Key Contributions and Results

A. Characterization of Macroscopic States

The paper provides four equivalent characterizations for a state $\rho$ to be "macroscopic" with respect to a POVM $P$ and prior $\gamma$:

1. Zero Observational Deficit: $\delta_P(\rho \| \gamma) = 0$.
2. Fixed Point of Coarse-Graining: $C_{P,\gamma}(\rho) = \rho$.
3. Fixed Point of Resource Destroying Map: $\Delta_{P,\gamma}(\rho) = \rho$, where $\Delta_{P,\gamma}$ is an idempotent CPTP map.
4. Algebraic Structure: $\rho$ can be written as $\sum_y c_y \Pi_y \gamma$, where $\Pi_y$ are elements of a specific Projection-Valued Measure (PVM).

B. Maximal Projective Post-Processing (MPPP)

A central theoretical result is the existence and uniqueness of the Maximal Projective Post-Processing (MPPP), denoted $\Pi_{P,\gamma}$.

• Definition: It is the unique PVM that is a classical post-processing of the POVM $P$ and whose elements commute with the prior $\gamma$.
• Significance: The MPPP acts as an "inferential reference frame." It encapsulates the minimal yet sufficient information required for optimal retrodiction. The set of macroscopic states corresponds exactly to the states that are block-diagonal with respect to this MPPP and have specific block structures determined by $\gamma$.

C. Resource Theory of Microscopicity

The authors establish a resource theory where:

• Free States: Macroscopic states (those with zero observational deficit).
• Resource Destroying Map (RDM): The map $\Delta_{P,\gamma}$, which projects any state onto the set of macroscopic states.
• Free Operations: Three hierarchically ordered classes are defined:
  1. Microscopicity Non-Generating Operations (MNO): Maps that preserve the set of macroscopic states.
  2. Resource-Destroying Covariant Operations (RCO): Operations commuting with the RDM.
  3. Coarse-Graining Covariant Operations (CCO): Operations commuting with the coarse-graining map $C_{P,\gamma}$.
  • Result: $CCO \subseteq RCO \subseteq MNO$.
• Unification: This framework generalizes and unifies existing resource theories, including Coherence (when $P$ is rank-1 and $\gamma$ is diagonal), Athermality (when the MPPP is trivial and $\gamma$ is a Gibbs state), and Asymmetry.

D. Observational Discord and Quantum Correlations

The paper introduces Observational Discord ($D_{P_A}$), a measure of quantum correlations accessible to a specific observer with limited measurement capabilities (POVM $P_A$).

• Definition: It is the difference between the total mutual information and the mutual information accessible after the local measurement $P_A$.
• Vanishing Condition: The authors derive necessary and sufficient conditions for observational discord to vanish. Unlike standard quantum discord (which vanishes only for classical-quantum states), observational discord vanishes if the state is a "locally macroscopic" state relative to the observer's specific $P_A$ and prior.
• Insight: This demonstrates that quantum correlations are context-dependent resources. Their visibility and utility are determined by the observer's inferential reference frame, not just the state's intrinsic properties.

4. Significance and Implications

• Foundational Clarity: The work provides a rigorous mathematical foundation for "macroscopicity," moving beyond heuristic definitions to a precise inferential framework grounded in Bayesian retrodiction and quantum information theory.
• Unification of Thermodynamics and Information: By linking observational entropy to the Petz recovery map and statistical sufficiency, the paper bridges the gap between thermodynamic irreversibility and information loss.
• Observer-Dependent Reality: It challenges the view of quantum correlations as absolute. Instead, it posits that the "quantumness" of a system is relative to the observer's measurement resolution and prior knowledge. This is crucial for realistic scenarios where observers have limited control or access.
• Generalization of Resource Theories: The "Resource Theory of Microscopicity" offers a powerful generalization that subsumes coherence, asymmetry, and athermality, suggesting a unified approach to studying quantum resources under various constraints.
• Future Directions: The paper opens avenues for studying entropy production in coarse-grained systems, catalytic transformations in macroscopic settings, and the limits of measurements imposed by symmetry (Wigner-Araki-Yanase theorem).

In summary, Nagasawa et al. present a comprehensive framework that redefines macroscopic states as those recoverable from macroscopic data, quantifies the cost of this irreversibility via observational deficit, and demonstrates how the perception of quantum correlations is fundamentally tied to the observer's inferential capabilities.