Implicit Representations of Meaning in Neural Language Models - ACL Anthology

Belinda Z. Li, Maxwell Nye, Jacob Andreas. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

1 mentions: @ELYZA_inc
Date: 2021/10/14 10:38

Referring Tweets

@ELYZA_inc 【Contextual embeddingは「意味」を獲得しているか?】 動的意味論の枠組みに則り、BARTやT5が学習したcontextual embeddingが、「文脈が記述する状況」に関する命題の真理値を正しく分類できるかをprobing。テキストデータのみから意味情報を獲得していることを示唆した。

Related Entries

Read more [2005.01795] Generating SOAP Notes from Doctor-Patient Conversations Using Modular Summarization Tec...
0 users, 1 mentions 2021/11/07 10:37
Read more [2109.02492v1] DialogLM: Pre-trained Model for Long Dialogue Understanding and Summarization
0 users, 1 mentions 2021/09/19 10:37
Read more [2109.06513] Exploring Prompt-based Few-shot Learning for Grounded Dialog Generation
0 users, 1 mentions 2021/10/31 10:37
Read more BabyBERTa: Learning More Grammar With Small-Scale Child-Directed Language - ACL Anthology
0 users, 1 mentions 2021/11/12 10:38
Read more Are Factuality Checkers Reliable? Adversarial Meta-evaluation of Factuality in Summarization - ACL A...
0 users, 1 mentions 2021/12/04 10:40



安定したサイト運営のためにGitHub sponsorを募集しています。