Towards credible visual model interpretation with path attribution
Originally inspired by game-theory, path attribution framework stands out among the post-hoc model interpretation tools due to its axiomatic nature. However, recent developments show that this framework can still suffer from counter-intuitive results. Moreover, specifically for deep visual models, the existing path-based methods also fall short on conforming to the original intuitions that are the basis of the claimed axiomatic properties of this framework. We address these problems with a systematic investigation, and pinpoint the conditions in which the counter-intuitive results can be avoided for deep visual model interpretation with the path attribution strategy. We also devise a scheme to preclude the conditions in which visual model interpretation can invalidate the axiomatic properties of path attribution. These insights are combined into a method that enables reliable visual model interpretation. Our findings are establish empirically with multiple datasets, models and evaluation metrics. Extensive experiments show a consistent performance gain of our method over the baselines.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Path Choice Matters for Clear Attribution in Path Methods
Rigorousness and clarity are both essential for interpretations of DNNs to engender human trust. Path methods are commonly employed to generate rigorous attributions that satisfy three axioms. However, the meaning of att…
Towards Semantic Interpretation of Thoracic Disease and COVID-19 Diagnosis Models
Convolutional neural networks are showing promise in the automatic diagnosis of thoracic pathologies on chest x-rays. Their black-box nature has sparked many recent works to explain the prediction via input feature attri…
COVID-19 DiagnosisPredictionNeuroWise: A Multi-Agent LLM "Glass-Box" System for Practicing Double-Empathy Communication with Autistic Partners
The double empathy problem frames communication difficulties between neurodivergent and neurotypical individuals as arising from mutual misunderstanding, yet most interventions focus on autistic individuals. We present N…
Beyond Single Path Integrated Gradients for Reliable Input Attribution via Randomized Path Sampling
Input attribution is a widely used explanation method for deep neural networks, especially in visual tasks. Among various attribution methods, Integrated Gradients (IG) is frequently used because of its model-agnosti…
Interpreting Interpretations: Organizing Attribution Methods by Criteria
Motivated by distinct, though related, criteria, a growing number of attribution methods have been developed tointerprete deep learning. While each relies on the interpretability of the concept of "importance" and our ab…
image-classificationImage Classification