First-hand information for everyone
A new theoretical study recasts a key moment in black hole evaporation — the so-called Page transition — as a change in the information-theoretic cost of competing semiclassical descriptions. Using a two-dimensional quantum gravity model, the authors develop a replica-trick framework to compute quantum relative entropy and apply it to compare the exact quantum radiation state with two familiar semiclassical pictures: the original Hawking saddle (a disconnected geometry) and the replica-wormhole saddle (a connected geometry).
Relative entropy is an information-theoretic measure of distinguishability between two quantum states (classically related to Kullback–Leibler divergence). In this work it plays the role of a diagnostic: how ‘‘expensive’’ it is, in information terms, to describe the true radiation state using a given coarse-grained semiclassical approximation. The paper reports two central, opposing trends as evaporation proceeds. The relative entropy between the exact radiation state and the Hawking saddle grows monotonically, meaning the Hawking description becomes steadily worse (or more costly) at capturing the true quantum state. By contrast, the relative entropy between the exact state and the replica-wormhole saddle decreases, indicating that the connected-geometry description becomes progressively more accurate.
Crucially, the sign of the difference between these two relative entropies flips at precisely the Page time — the moment when the radiation’s von Neumann entropy peaks and begins to follow the so-called Page curve. This flip provides an information-theoretic characterization of the Page transition: before the Page time, the disconnected Hawking saddle is the cheaper coarse-grained description; after the Page time, the replica-wormhole saddle is preferred. The authors interpret this as a flow of relative-entropy costs between competing semiclassical descriptions and argue that this perspective illuminates the emergence of effective irreversibility (coarse-grained loss of information) in black hole evaporation.
The work uses Jackiw–Teitelboim (JT) gravity, a two-dimensional model that has become a standard testing ground for conceptual issues in quantum gravity because it is tractable while retaining features thought to be relevant for higher-dimensional black holes. Technically, the authors extend replica-trick techniques — a method that computes entropic quantities by studying n-fold replicated systems and analytically continuing n to 1 — to the calculation of relative entropy. This allows explicit, controlled comparisons of the competing saddles’ information-theoretic costs.
The findings are conceptually notable because they shift the Page transition from being described solely as a change in entanglement entropy to being understood as a crossing of relative-entropy costs between approximations. That reframing helps quantify when and why a semiclassical description breaks down, and it connects gravitational saddle competition to operational notions from quantum information theory.
Caveats are important. JT gravity is a simplified toy model, and replica-trick computations require subtle analytic continuations and assumptions about dominant saddles. The exact radiation state used for comparison is available in these controlled models but is not accessible for realistic astrophysical black holes. Thus, while the results provide a clean conceptual diagnosis and a useful calculational framework, they do not by themselves resolve all aspects of the black hole information problem in higher dimensions.
Looking ahead, the framework suggests several directions: extending the analysis beyond two dimensions, testing robustness against different matter content or interactions, and exploring connections to quantum error-correcting descriptions of spacetime. Even if limited to toy models for now, casting the Page transition as a flow of relative-entropy costs offers a fresh, quantitative vocabulary for thinking about how semiclassical gravity emerges and when it must be replaced by a more complete quantum description.