Generative Example-Based Explanations: Bridging the Gap between Generative Modeling and Explainability

Recently, several methods have leveraged deep generative modeling to produce example-based explanations of decision algorithms for high-dimensional input data. Despite promising results, a disconnect exists between these methods and the classical explainability literature, which focuses on lower-dimensional data with semantically meaningful features. This conceptual and communication gap leads to misunderstandings and misalignments in goals and expectations. In this paper, we bridge this gap by proposing a novel probabilistic framework for local example-based explanations. Our framework integrates the critical characteristics of classical local explanation desiderata while being amenable to high-dimensional data and their modeling through deep generative models. Our aim is to facilitate communication, foster rigor and transparency, and improve the quality of peer discussion and research progress.

GradCheck: Analyzing classifier guidance gradients for conditional diffusion sampling

To sample from an unconditionally trained Denoising Diffusion Probabilistic Model (DDPM), classifier guidance adds conditional information during sampling, but the gradients from classifiers, especially those not trained on noisy images, are often unstable. This study conducts a gradient analysis comparing robust and non-robust classifiers, as well as multiple gradient stabilization techniques. Experimental results demonstrate that these techniques significantly improve the quality of class-conditional samples for non-robust classifiers by providing more stable and informative classifier guidance gradients. The findings highlight the importance of gradient stability in enhancing the performance of classifier guidance, especially on non-robust classifiers.

Diffusion-based Visual Counterfactual Explanations -- Towards Systematic Quantitative Evaluation

Latest methods for visual counterfactual explanations (VCE) harness the power of deep generative models to synthesize new examples of high-dimensional images of impressive quality. However, it is currently difficult to compare the performance of these VCE methods as the evaluation procedures largely vary and often boil down to visual inspection of individual examples and small scale user studies. In this work, we propose a framework for systematic, quantitative evaluation of the VCE methods and a minimal set of metrics to be used. We use this framework to explore the effects of certain crucial design choices in the latest diffusion-based generative models for VCEs of natural image classification (ImageNet). We conduct a battery of ablation-like experiments, generating thousands of VCEs for a suite of classifiers of various complexity, accuracy and robustness. Our findings suggest multiple directions for future advancements and improvements of VCE methods. By sharing our methodology and our approach to tackle the computational challenges of such a study on a limited hardware setup (including the complete code base), we offer a valuable guidance for researchers in the field fostering consistency and transparency in the assessment of counterfactual explanations.