Time-Dependent Quantum Monte Carlo for fermions: from Bayesian conditioning to spinor dynamics
The time-dependent quantum Monte Carlo method represents many-electron states using an ensemble of replicas where electrons are tracked by walkers sampling physical space. The paper derives a conditional interaction from Bayesian theorem, yielding a meaningful resolution length. It extends the method to spinors, allowing it to capture non-collinear spin dynamics in systems like double quantum dots, recovering entanglement missed by simpler models.