@inproceedings{alvarez-melis-jaakkola-2017-causal,
title = "A causal framework for explaining the predictions of black-box sequence-to-sequence models",
author = "Alvarez-Melis, David and
Jaakkola, Tommi",
editor = "Palmer, Martha and
Hwa, Rebecca and
Riedel, Sebastian",
booktitle = "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
month = sep,
year = "2017",
address = "Copenhagen, Denmark",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/D17-1042",
doi = "10.18653/v1/D17-1042",
pages = "412--421",
abstract = "We interpret the predictions of any black-box 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 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-to-sequence problems, adopting a variational autoencoder to yield meaningful input perturbations. We test our method across several NLP sequence generation tasks.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="alvarez-melis-jaakkola-2017-causal">
<titleInfo>
<title>A causal framework for explaining the predictions of black-box sequence-to-sequence models</title>
</titleInfo>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Alvarez-Melis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tommi</namePart>
<namePart type="family">Jaakkola</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2017-09</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Martha</namePart>
<namePart type="family">Palmer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rebecca</namePart>
<namePart type="family">Hwa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sebastian</namePart>
<namePart type="family">Riedel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Copenhagen, Denmark</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We interpret the predictions of any black-box 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 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-to-sequence problems, adopting a variational autoencoder to yield meaningful input perturbations. We test our method across several NLP sequence generation tasks.</abstract>
<identifier type="citekey">alvarez-melis-jaakkola-2017-causal</identifier>
<identifier type="doi">10.18653/v1/D17-1042</identifier>
<location>
<url>https://aclanthology.org/D17-1042</url>
</location>
<part>
<date>2017-09</date>
<extent unit="page">
<start>412</start>
<end>421</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T A causal framework for explaining the predictions of black-box sequence-to-sequence models
%A Alvarez-Melis, David
%A Jaakkola, Tommi
%Y Palmer, Martha
%Y Hwa, Rebecca
%Y Riedel, Sebastian
%S Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing
%D 2017
%8 September
%I Association for Computational Linguistics
%C Copenhagen, Denmark
%F alvarez-melis-jaakkola-2017-causal
%X We interpret the predictions of any black-box 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 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-to-sequence problems, adopting a variational autoencoder to yield meaningful input perturbations. We test our method across several NLP sequence generation tasks.
%R 10.18653/v1/D17-1042
%U https://aclanthology.org/D17-1042
%U https://doi.org/10.18653/v1/D17-1042
%P 412-421
Markdown (Informal)
[A causal framework for explaining the predictions of black-box sequence-to-sequence models](https://aclanthology.org/D17-1042) (Alvarez-Melis & Jaakkola, EMNLP 2017)
ACL