pith. sign in

To achieve this faithfully, training must be done in eval mode, i.e., modules like batch norm and dropout should not be active

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

stat.ML 1

years

2025 1

verdicts

UNVERDICTED 1

representative citing papers

TPV: Parameter Perturbations Through the Lens of Test Prediction Variance

stat.ML · 2025-12-11 · unverdicted · novelty 7.0

TPV measures first-order sensitivity of model outputs to parameter perturbations, unifies robustness analysis under one lens, proves train-to-test convergence in overparameterized limits, and enables label-free pruning and model selection applications.

citing papers explorer

Showing 1 of 1 citing paper.

  • TPV: Parameter Perturbations Through the Lens of Test Prediction Variance stat.ML · 2025-12-11 · unverdicted · none · ref 8

    TPV measures first-order sensitivity of model outputs to parameter perturbations, unifies robustness analysis under one lens, proves train-to-test convergence in overparameterized limits, and enables label-free pruning and model selection applications.