Shortcut Learning of Large Language Models in Natural Language Understanding.
In: Communications of the ACM, Jg. 67 (2024), Heft 1, S. 110-120
serialPeriodical
Zugriff:
The article looks at the use of large language models to carry out natural language understanding (NLU) tasks. It suggests that the shortcut learning common to existing large language models based on machine learning limits how robust their performance can be because they are overly dependent on spurious correlations and incidental relationships. It discusses possible approaches to overcoming this problem in the future development of large language models.
Titel: |
Shortcut Learning of Large Language Models in Natural Language Understanding.
|
---|---|
Autor/in / Beteiligte Person: | MENGNAN, DU ; FENGXIANG, HE ; NA, ZOU ; DACHENG, TAO ; XIA, HU |
Zeitschrift: | Communications of the ACM, Jg. 67 (2024), Heft 1, S. 110-120 |
Veröffentlichung: | 2024 |
Medientyp: | serialPeriodical |
ISSN: | 0001-0782 (print) |
DOI: | 10.1145/3596490 |
Schlagwort: |
|
Sonstiges: |
|