Universal crossovers in weakly-monitored quantum critical states
arXiv:2608.02716
2026
Dynamics
2 ideas extracted · analyzed Aug 31, 2026
What the math gives to ML
The paper presents a transferable renormalization-group mechanism: weak, unpostselected measurement noise is a relevant perturbation that grows under coarse graining and drives a crossover from clean critical behavior to a measurement-dominated fixed point. The crossover scale is controlled by the perturbation strength and its RG eigenvalue, with distinct long-scale regimes including area-law and logarithmic-entanglement behavior. In neural networks, this can motivate stochastic activation or attention perturbations whose strength is scheduled against effective depth, sequence length, or training time. The strongest engineering tests are scaling-collapse experiments and controllers based on moment growth, rather than benchmark improvements alone.
Ideas from this paper
✗ Failed on benchmark
2026
Inject weak, unpostselected stochastic perturbations into activations, attention links, or recurrent transitions, but scale their strength according to effective computational size. The schedule is designed so that noise is initially a weak perturbation and becomes dominant only beyond a controlled depth or sequence length, producing a measurable crossover rather than uncalibrated constant dropout.
Useful7/10
Difficulty4/10
Novelty6/10
Unverified
2026
Monitor moments of the network's response to independent stochastic forward passes instead of tracking only mean loss or mean activation variance. Nonlinear moment scaling detects intermittent and heterogeneous sensitivity, allowing a controller to reduce noise or learning rate before average metrics reveal instability.
Useful6/10
Difficulty5/10
Novelty7/10