Large-N dynamics of the spiked tensor model with random initial conditions
arXiv:2208.12586v1h-index: 6
Originality Synthesis-oriented
AI Analysis
This work addresses theoretical physics problems in statistical mechanics and random matrix theory, but it appears incremental as it extends existing methods to a specific model.
The authors tackled the dynamics of the spiked tensor model with random initial conditions by developing a path integral approach for partial differential equations, showing that large-N saddle point equations are dominated by melonic diagrams.
In these notes, we develop a path integral approach for the partial differential equations with random initial conditions. Then, we apply it to the dynamics of the spiked tensor model and show that the large-$N$ saddle point equations are dominated by the melonic type diagrams.