Language Model Behavior: A Comprehensive Survey
Computational Linguistics · 01.03.2024 · 50 (1) · 293-350
DOI: 10.1162/coli_a_00492 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computational Linguistics, online erschienen am 01.03.2024. Verfasst von Tyler A. Chang und 1 weiteren Autor:innen. Die Arbeit wurde bisher 100-mal zitiert (Crossref) und nennt 393 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Transformer language models have received widespread public attention, yet their generated text is often surprising even to NLP researchers. In this survey, we discuss over 250 recent studies of English language model behavior before task-specific fine-tuning. Language models possess basic capabilities in syntax, semantics, pragmatics, world knowledge, and reasoning, but these capabilities are sensitive to specific inputs and surface features. Despite dramatic increases in generated text quality as models scale to hundreds of billions of parameters, the models are still prone to unfactual responses, commonsense errors, memorized text, and social biases. Many of these weaknesses can be framed as over-generalizations or under-generalizations of learned patterns in text. We synthesize recent results to highlight what is currently known about large language model capabilities, thus providing a resource for applied work and for research in adjacent fields that use language models.
Zitieren
APA
Chang, T. A., & Bergen, B. K. (2024). Language Model Behavior: A Comprehensive Survey. Computational Linguistics, 50(1), 293-350. https://doi.org/10.1162/coli_a_00492
Vancouver
Chang TA, Bergen BK. Language Model Behavior: A Comprehensive Survey. Computational Linguistics. 2024;50(1):293-350. doi:10.1162/coli_a_00492
BibTeX
@article{Chang2024,
author = {Chang, Tyler A. and Bergen, Benjamin K.},
title = {Language Model Behavior: A Comprehensive Survey},
journal = {Computational Linguistics},
year = {2024},
volume = {50},
number = {1},
pages = {293--350},
doi = {10.1162/coli_a_00492}
}
EUVIMED