SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.09890 · ACL · 2020

Considering Likelihood in NLP Classification Explanations with Occlusion and Language Modeling

Christoph Alt, David Harbecke

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Recently, state-of-the-art NLP models gained an increasing syntactic and semantic understanding of language, and explanation methods are crucial to understand their decisions. Occlusion is a well established method that provides explanations on discrete language data, e.g. by removing a language unit from an input and measuring the impact on a model's decision. We argue that current occlusionbased methods often produce invalid or syntactically incorrect language data, neglecting the improved abilities of recent NLP models. Furthermore, gradient-based explanation methods disregard the discrete distribution of data in NLP. Thus, we propose OLM: a novel explanation method that combines occlusion and language models to sample valid and syntactically correct replacements with high likelihood, given the context of the original input. We lay out a theoretical foundation that alleviates these weaknesses of other explanation methods in NLP and provide results that underline the importance of considering data likelihood in occlusion-based explanation. 1

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2004.09890")
get_code_for_paper("2004.09890")
have("2004.09890")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2004.09890.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2004.09890.svg)](https://syntology.ai/paper/2004.09890)

Connect an agent — have() is free.