We identify test prediction variance (TPV) -- the first-order sensitivity of model outputs to parameter perturbations around a trained solution -- as a unifying quantity that links several classical observations about generalization in deep networks. TPV is a fully label-free object whose trace form separates the geometry of the trained model from the specific perturbation mechanism, allowing a broad family of parameter perturbations like SGD noise, label noise, finite-precision noise, and other post-training perturbations to be analyzed under a single framework. Theoretically, we show that TPV estimated on the training set converges to its test-set value in the overparameterized limit, providing the first result that prediction variance under local parameter perturbations can be inferred from training inputs alone. Empirically, TPV exhibits a striking stability across datasets and architectures -- including extremely narrow networks -- and correlates well with clean test loss. Finally, we demonstrate that modeling pruning as a TPV perturbation yields a simple label-free importance measure that performs competitively with state-of-the-art pruning methods, illustrating the practical utility of TPV. Code available at github.com/devansharpit/TPV.

Self-paced learning and hard example mining re-weight training instances to improve learning accuracy.

Neural network robustness has become a central topic in machine learning in recent years. Most training algorithms that improve the model's robustness to

Neural network robustness has become a central topic in machine learning in recent years. Most training algorithms that improve the model's robustness to

We propose an active inference agent to identify and control a mechanical system with multiple bodies connected by joints. This agent is constructed from multiple scalar autoregressive model-based agents, coupled together by virtue of sharing memories. Each subagent infers parameters through Bayesian filtering and controls by minimizing expected free energy over a finite time horizon. We demonstrate that a coupled agent of this kind is able to learn the dynamics of a double mass-spring-damper system, and drive it to a desired position through a balance of explorative and exploitative actions. It outperforms the uncoupled subagents in terms of surprise and goal alignment.

Self-paced learning and hard example mining re-weight training instances to improve learning accuracy. This paper presents two improved alternatives based on lightweight estimates of sample uncertainty in stochastic gradient descent (SGD): the variance in predicted probability of the correct class across iterations of minibatch SGD, and the proximity of the correct class probability to the decision threshold. Extensive experimental results on six datasets show that our methods reliably improve accuracy in various network architectures, including additional gains on top of other popular training techniques, such as residual learning, momentum, ADAM, batch normalization, dropout, and distillation.

Graph Neural Networks (GNNs) have achieved promising results in tasks such as node classification and graph classification. However, recent studies reveal that GNNs are vulnerable to backdoor attacks, posing a significant threat to their real-world adoption. Despite initial efforts to defend against specific graph backdoor attacks, there is no work on defending against various types of backdoor attacks where generated triggers have different properties. Hence, we first empirically verify that prediction variance under edge dropping is a crucial indicator for identifying poisoned nodes. With this observation, we propose using random edge dropping to detect backdoors and theoretically show that it can efficiently distinguish poisoned nodes from clean ones. Furthermore, we introduce a novel robust training strategy to efficiently counteract the impact of the triggers. Extensive experiments on real-world datasets show that our framework can effectively identify poisoned nodes, significantly degrade the attack success rate, and maintain clean accuracy when defending against various types of graph backdoor attacks with different properties.

In healthcare, thanks to many model agnostic methods, explainability of the prediction scores made by deep learning applications has improved. However, we note that for daily or hourly risk of deterioration prediction of in-hospital patients, not only the predicted risk probability score matters, but also the variance of the risk scores play key roles in aiding clinical decision making. In this paper, we propose to use delta's method to approximate variance of prediction deterministically, such that the SHAP method can be adopted to attribute contribution of variance. The prediction variance is estimated by sampling the conditional hidden space in variational models and is propagated to input clinical variables based on Shapley values of the variance game. This approach works with variational time series models such as variational recurrent neural networks and variational transformers. We further argue that variational time series models are perfect fits for achieving a balance between predictive power and explainability through a series of experiments on a public clinical ICU datasets. Since SHAP values are additive, we also postulate that the SHAP importance of clinical variables with respect to prediction variations can guide their frequency of measurements.

Modeling and prediction of epidemic spread are critical to assist in policy-making for mitigation. Therefore, we present a new method based on Gaussian Process Regression to model and predict epidemics, and it quantifies prediction confidence through variance and high probability error bounds. Gaussian Process Regression excels in using small datasets and providing uncertainty bounds, and both of these properties are critical in modeling and predicting epidemic spreading processes with limited data. However, the derivation of formal uncertainty bounds remains lacking when using Gaussian Process Regression in the setting of epidemics, which limits its usefulness in guiding mitigation efforts. Therefore, in this work, we develop a novel bound on the variance of the prediction that quantifies the impact of the epidemic data on the predictions we make. Further, we develop a high probability error bound on the prediction, and we quantify how the epidemic spread, the infection data, and the length of the prediction horizon all affect this error bound. We also show that the error stays below a certain threshold based on the length of the prediction horizon. To illustrate this framework, we leverage Gaussian Process Regression to model and predict COVID-19 using real-world infection data from the United Kingdom.

Model stealing aims at inferring a victim model's functionality at a fraction of the original training cost. While the goal is clear, in practice the model's architecture, weight dimension, and original training data can not be determined exactly, leading to mutual uncertainty during stealing. In this work, we explicitly tackle this uncertainty by generating multiple possible networks and combining their predictions to improve the quality of the stolen model. For this, we compare five popular uncertainty quantification models in a model stealing task. Surprisingly, our results indicate that the considered models only lead to marginal improvements in terms of label agreement (i.e., fidelity) to the stolen model. To find the cause of this, we inspect the diversity of the model's prediction by looking at the prediction variance as a function of training iterations. We realize that during training, the models tend to have similar predictions, indicating that the network diversity we wanted to leverage using uncertainty quantification models is not (high) enough for improvements on the model stealing task.