Visualizing the diversity of representations learned by Bayesian neural networks

Grinwald, Dennis; Bykov, Kirill; Nakajima, Shinichi; Höhne, Marina M. -C.

doi:10.48550/arxiv.2201.10859

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Intuitively, in contrast to the MAP learning, where point estimates of weights represent one deterministic decision-making strategy, a posterior distribution represents an infinite ensemble of models, which employ different strategies towards the prediction. By aggregating the variability of the decision-making processes of networks, we can obtain a broader outlook on the features that were used for the prediction, and thus deeper insights into the models' behavior (Grinwald et al 2022).…”

Section: Introductionmentioning

confidence: 99%

NoiseGrad — Enhancing Explanations by Introducing Stochasticity to Model Weights

Bykov¹,

Hedström

Nakajima

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

Many efforts have been made for revealing the decision-making process of black-box learning machines such as deep neural networks, resulting in useful local and global explanation methods. For local explanation, stochasticity is known to help: a simple method, called SmoothGrad, has improved the visual quality of gradient-based attribution by adding noise to the input space and averaging the explanations of the noisy inputs. In this paper, we extend this idea and propose NoiseGrad that enhances both local and global explanation methods. Specifically, NoiseGrad introduces stochasticity in the weight parameter space, such that the decision boundary is perturbed. NoiseGrad is expected to enhance the local explanation, similarly to SmoothGrad, due to the dual relationship between the input perturbation and the decision boundary perturbation. We evaluate NoiseGrad and its fusion with SmoothGrad - FusionGrad - qualitatively and quantitatively with several evaluation criteria, and show that our novel approach significantly outperforms the baseline methods. Both NoiseGrad and FusionGrad are method-agnostic and as handy as SmoothGrad using a simple heuristic for the choice of the hyperparameter setting without the need of fine-tuning.

show abstract

Section: Introductionmentioning

confidence: 99%