Transformers reconstruct the constituent RCFTs in tensor-product theories from low-energy spectra, reaching 98% accuracy on WZW models and generalizing to larger central charges with few out-of-domain examples.
Decoding conformal field theories: from supervised to unsupervised learning
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
hep-th 2years
2026 2roles
background 1polarities
background 1representative citing papers
Numerical search finds candidate modular-invariant spectra with integer degeneracies for 1 < c < 8/7 and hints at a stronger gap bound near c = 1.
citing papers explorer
-
Reconstructing conformal field theoretical compositions with Transformers
Transformers reconstruct the constituent RCFTs in tensor-product theories from low-energy spectra, reaching 98% accuracy on WZW models and generalizing to larger central charges with few out-of-domain examples.
-
Descending into the Modular Bootstrap
Numerical search finds candidate modular-invariant spectra with integer degeneracies for 1 < c < 8/7 and hints at a stronger gap bound near c = 1.