Work

A causal framework for explaining the predictions of black-box sequence-to-sequence models

Year: 2017

Type: conference-paper

Abstract: We interpret the predictions of any blackbox structured input-structured output model around a specific input-output pair.Our method returns an "explanation" consisting of groups of input-output tokens that are causally related.These dependencies are inferred by querying the black-box model with perturbed inputs, generating a graph over tokens from the responses, and solving a partitioning problem to select the most relevant components.We focus the general approach on sequence-tosequence problems, adopting a variational autoencoder to yield meaningful input perturbations.We test our method across several NLP sequence generation tasks.

Source: Conference on Empirical Methods in Natural Language Processing (EMNLP)

DOI: 10.18653/v1/d17-1042

Authors: David Alvarez-Melis, Tommi Jaakkola

Institutions: MIT Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology

Language: English

Topic: Explainable Artificial Intelligence (XAI)

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funder: Consejo Nacional de Ciencia y Tecnología

OpenAlex ID: W2724701856