TY - GEN
T1 - Entailment Semantics Can Be Extracted from an Ideal Language Model
AU - Merrill, William
AU - Warstadt, Alex
AU - Linzen, Tal
N1 - Publisher Copyright:
©2022 Association for Computational Linguistics.
PY - 2022
Y1 - 2022
N2 - Language models are often trained on text alone, without additional grounding. There is debate as to how much of natural language semantics can be inferred from such a procedure. We prove that entailment judgments between sentences can be extracted from an ideal language model that has perfectly learned its target distribution, assuming the training sentences are generated by Gricean agents, i.e., agents who follow fundamental principles of communication from the linguistic theory of pragmatics. We also show entailment judgments can be decoded from the predictions of a language model trained on such Gricean data. Our results reveal a pathway for understanding the semantic information encoded in unlabeled linguistic data and a potential framework for extracting semantics from language models.
AB - Language models are often trained on text alone, without additional grounding. There is debate as to how much of natural language semantics can be inferred from such a procedure. We prove that entailment judgments between sentences can be extracted from an ideal language model that has perfectly learned its target distribution, assuming the training sentences are generated by Gricean agents, i.e., agents who follow fundamental principles of communication from the linguistic theory of pragmatics. We also show entailment judgments can be decoded from the predictions of a language model trained on such Gricean data. Our results reveal a pathway for understanding the semantic information encoded in unlabeled linguistic data and a potential framework for extracting semantics from language models.
UR - http://www.scopus.com/inward/record.url?scp=85153313157&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85153313157&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85153313157
T3 - CoNLL 2022 - 26th Conference on Computational Natural Language Learning, Proceedings of the Conference
SP - 176
EP - 193
BT - CoNLL 2022 - 26th Conference on Computational Natural Language Learning, Proceedings of the Conference
PB - Association for Computational Linguistics (ACL)
T2 - 26th Conference on Computational Natural Language Learning, CoNLL 2022 collocated and co-organized with EMNLP 2022
Y2 - 7 December 2022 through 8 December 2022
ER -