Deep learning emergent spacetime from fermionic spectral functions in holography
Abstract
We present a physics-informed machine learning framework based on Neural Ordinary Differential Equations that solves the holographic inverse problem: reconstructing the bulk spacetime and gauge field of a charged AdS black hole directly from boundary fermionic spectral functions. Encoding the UV asymptotics, horizon regularity, and zero temperature extremality as hard constraints in the neural network architecture, our framework reliably reconstructs the extremal Reissner-Nordström AdS geometry across three quantum critical regimes set by the U(1) probe charge---non-Fermi liquid, marginal Fermi liquid (strange metal), and Fermi-liquid-like states---and can jointly infer the probe charge itself to sub-percent accuracy. Relaxing the near-AdS boundary constraint uncovers a geometrical degeneracy: bulk profiles that differ throughout the radial direction but share the same near-horizon AdS2 × ℝ2 data reproduce identical spectral functions near the Fermi surface. This isospectral non-uniqueness is precisely the bulk degeneracy expected on general holographic grounds at zero temperature, and its spontaneous emergence across independent training runs shows that the network isolates the IR CFT universality rather than overfitting a single UV completion.
Related
- dS/CFT Holography including Fermionic Fields
- Fermionic Signatures of Antispacetime-spacetime Domain Walls
- Fermionic Particle Creation in Curved Spacetime: Physical Bloch-Vector Dynamics and Exact de Sitter Scaling
- Fermionic Spectral Functions in a Two-Current Gubser-Rocha Model with Axion Momentum Relaxation
- Symmetry-driven correlation patterns in one-dimensional periodic fermionic systems: closed-form expressions and exact selection rules for correlation functions, entanglement entropies and mutual information
BibTeX
@misc{hashimoto2026deep,
title = {{Deep learning emergent spacetime from fermionic spectral functions in holography}},
author = {Koji Hashimoto and Hyun-Sik Jeong and Keun-Young Kim and Daichi Takeda and Kwan Yun},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.18566},
eprint = {2609.18566},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.18566},
}
RIS
TY - UNPB
TI - Deep learning emergent spacetime from fermionic spectral functions in holography
AU - Koji Hashimoto
AU - Hyun-Sik Jeong
AU - Keun-Young Kim
AU - Daichi Takeda
AU - Kwan Yun
PY - 2026
DA - 2026/09/16
PB - arXiv
DO - 10.48550/arxiv.2609.18566
UR - https://arxiv.org/abs/2609.18566
AB - We present a physics-informed machine learning framework based on Neural Ordinary Differential Equations that solves the holographic inverse problem: reconstructing the bulk spacetime and gauge field of a charged AdS black hole directly from boundary fermionic spectral functions. Encoding the UV asymptotics, horizon regularity, and zero temperature extremality as hard constraints in the neural network architecture, our framework reliably reconstructs the extremal Reissner-Nordström AdS geometry across three quantum critical regimes set by the U(1) probe charge---non-Fermi liquid, marginal Fermi liquid (strange metal), and Fermi-liquid-like states---and can jointly infer the probe charge itself to sub-percent accuracy. Relaxing the near-AdS boundary constraint uncovers a geometrical degeneracy: bulk profiles that differ throughout the radial direction but share the same near-horizon AdS_2 × ℝ^2 data reproduce identical spectral functions near the Fermi surface. This isospectral non-uniqueness is precisely the bulk degeneracy expected on general holographic grounds at zero temperature, and its spontaneous emergence across independent training runs shows that the network isolates the IR CFT universality rather than overfitting a single UV completion.
ER -