Sound Research WIKINDX
Sound Research WIKINDX |
|
 |
|
Bender, E. M., & Koller, A. 2020, July. Climbing towards NLU: On meaning, form, and understanding in the age of data. Paper presented at Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.
Added by: alexb44 29/09/2026, 08:07 Last edited by: alexb44 29/09/2026, 08:12 |
|
| Resource type: Proceedings Article Published DOI: 10.18653/v1/2020.acl-main.463 BibTeX citation key: Bender2020 Email resource to friend View all bibliographic details  |
Categories: General Keywords: Language, Linguistics Creators: Bender, Chai, Jurafsky, Koller, Schluter, Tetreault Publisher: Association for Computational Linguistics Collection: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics |
Views: 3/6
|
| Abstract |
|
The success of the large neural language models on many NLP tasks is exciting. However, we find that these successes sometimes lead to hype in which these models are being described as ``understanding'' language or capturing ``meaning''. In this position paper, we argue that a system trained only on form has a priori no way to learn meaning. In keeping with the ACL 2020 theme of ``Taking Stock of Where We{'}ve Been and Where We{'}re Going'', we argue that a clear understanding of the distinction between form and meaning will help guide the field towards better science around natural language understanding.
Added by: alexb44 Last edited by: alexb44 |
| Notes |
|
Summary:
Grounds the hype (of 2020) of NLP applications by looking at meaning, form and understanding, building on Searle (Chinese Room) and Harnad (1990) (Symbol Grounding Problem). Devises octopus test to show this. They ultimately ask: "Are we climbing the right hill" (which I think is very pertinent to the development of LMs, still..). Form =/= meaning Added by: alexb44 Last edited by: alexb44 |