Neural networks reconstruct spectral functions in Minkowski QED

Researchers have utilized neural networks to reconstruct spectral functions in Minkowski quantum electrodynamics (QED), drawing inspiration from the dispersive formulation. The study involved solving the quenched rainbow QED Dyson-Schwinger equation and comparing results from neural reconstructions with free outputs against those with positivity-constrained ansätze. While free output neural networks successfully reproduced the expected zero crossing above the critical region, positivity-constrained models failed in the supercritical regime, suggesting spectral positivity should be a diagnostic rather than a blind constraint. AI

IMPACT Demonstrates a new method for applying neural networks to complex physics problems, potentially advancing scientific discovery.

RANK_REASON The cluster contains an academic paper detailing a novel application of neural networks to a physics problem. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Dyson--Schwinger equation
• Fukuda--Kugo equation
• Minkowski quantum electrodynamics
• Rodrigo Carmo Terin

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →