Deep Learning Metric Detectors in General Relativity

TL;DR

This paper explores the use of deep learning for detecting and comparing metrics in curved spacetimes, emphasizing coordinate invariance, and demonstrates applications in AdS spacetimes to estimate cosmological parameters.

Contribution

It introduces the concept of DL isometry for metric comparison, investigates its theoretical properties, and applies DL detectors to AdS spacetimes for parameter estimation.

Findings

DL isometry can identify equivalent metrics under coordinate transformations.

DL detectors can estimate cosmological constants in AdS spacetimes.

Superposition states may remain unchanged after DL measurement in quantum gravity context.

Abstract

We consider conceptual issues of deep learning (DL) for metric detectors using test particle geodesics in curved spacetimes. Advantages of DL metric detectors are emphasized from a view point of general coordinate transformations. Two given metrics (two spacetimes) are defined to be conneted by a DL isometry if their geodesic image data cannot be discriminated by any DL metric detector at any time. The fundamental question of when the DL isometry appears is extensively explored. If the two spacetimes connected by the DL isometry are in superposition of quantum gravity theory, the post-measurement state may be still in the same superposition even after DL metric detectors observe the superposed state. We also demonstrate metric-detection DL's in 2+1 dimensional anti-de Sitter (AdS) spacetimes to estimate the cosmological constants and Brown-Henneaux charges. In the AdS/CFT correspondence dictionary, it may be expected that such metric detectors in the AdS bulk region correspond to quantum measurement devices in the CFT at the AdS boundary.