Template Filling for Controllable Commonsense Reasoning

Dheeraj Rajagopal, Vivek Khetan, Bogdan Sacaleanu, Anatole Gershman, Andrew Fano, Eduard Hovy

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Large-scale sequence-to-sequence models have shown to be adept at both multiple-choice and open-domain commonsense reasoning tasks. However, the current formulations do not provide the ability to control the various attributes of the reasoning chain. To enable better controllability, we propose to study the commonsense reasoning as a template filling task (TemplateCSR) - where the language models fills reasoning templates with the given constraints as control factors. As an approach to TemplateCSR, we (i) propose a dataset of commonsense reasoning template-expansion pairs for healthcare and well-being domain and (ii) introduce ITO, an instruction fine-tuned sequence-to-sequence model that performs commonsense reasoning across concepts in the template. Our experiments show that our approach outperforms baseline both in generation metrics and factuality metrics. We also present a detailed error analysis on our approach's ability to reliably perform template based commonsense reasoning.

Original languageEnglish (US)
Title of host publicationIJCNLP-AACL 2023 - 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics, Findings of the Association for Computational Linguistics
Subtitle of host publicationIJCNLP-AACL 2023
EditorsJong C. Park, Yuki Arase, Baotian Hu, Wei Lu, Derry Wijaya, Ayu Purwarianti, Adila Alfa Krisnadhi
PublisherAssociation for Computational Linguistics (ACL)
Pages250-260
Number of pages11
ISBN (Electronic)9798891760189
StatePublished - 2023
Externally publishedYes
Event13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics: Findings of the Association for Computational Linguistic, IJCNLP-AACL 2023 - Nusa Dua, Bali, Indonesia
Duration: Nov 1 2023Nov 4 2023

Publication series

NameIJCNLP-AACL 2023 - 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics, Findings of the Association for Computational Linguistics: IJCNLP-AACL 2023

Conference

Conference13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics: Findings of the Association for Computational Linguistic, IJCNLP-AACL 2023
Country/TerritoryIndonesia
CityNusa Dua, Bali
Period11/1/2311/4/23

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Science Applications
  • Information Systems

Fingerprint

Dive into the research topics of 'Template Filling for Controllable Commonsense Reasoning'. Together they form a unique fingerprint.

Cite this