Sökning: WFRF:(Nivre Joakim 1962 )
> (2020-2024) >
Schrödinger's tree :
Schrödinger's tree : On syntax and neural language models
-
- Kulmizev, Artur (författare)
- Uppsala universitet,Institutionen för lingvistik och filologi,Uppsala University, Sweden
-
- Nivre, Joakim, 1962- (författare)
- RISE,Uppsala universitet,Institutionen för lingvistik och filologi,RISE Research Institutes of Sweden, Kista, Sweden,Datavetenskap,Uppsala University, Sweden
-
(creator_code:org_t)
- 2022-10-17
- 2022
- Engelska.
-
Ingår i: Frontiers in Artificial Intelligence. - : Frontiers Media S.A.. - 2624-8212. ; 5
- Relaterad länk:
-
https://doi.org/10.3...
-
visa fler...
-
https://uu.diva-port... (primary) (Raw object)
-
https://urn.kb.se/re...
-
https://doi.org/10.3...
-
https://urn.kb.se/re...
-
visa färre...
Abstract
Ämnesord
Stäng
- In the last half-decade, the field of natural language processing (NLP) hasundergone two major transitions: the switch to neural networks as the primarymodeling paradigm and the homogenization of the training regime (pre-train, then fine-tune). Amidst this process, language models have emergedas NLP’s workhorse, displaying increasingly fluent generation capabilities andproving to be an indispensable means of knowledge transfer downstream.Due to the otherwise opaque, black-box nature of such models, researchershave employed aspects of linguistic theory in order to characterize theirbehavior. Questions central to syntax—the study of the hierarchical structureof language—have factored heavily into such work, shedding invaluableinsights about models’ inherent biases and their ability to make human-likegeneralizations. In this paper, we attempt to take stock of this growing body ofliterature. In doing so, we observe a lack of clarity across numerous dimensions,which influences the hypotheses that researchers form, as well as theconclusions they draw from their findings. To remedy this, we urge researchersto make careful considerations when investigating coding properties, selectingrepresentations, and evaluating via downstream tasks. Furthermore, we outlinethe implications of the different types of research questions exhibited in studieson syntax, as well as the inherent pitfalls of aggregate metrics. Ultimately, wehope that our discussion adds nuance to the prospect of studying languagemodels and paves the way for a less monolithic perspective on syntax in thiscontext.
Ämnesord
- NATURVETENSKAP -- Data- och informationsvetenskap -- Språkteknologi (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Language Technology (hsv//eng)
Nyckelord
- neural networks
- language models
- syntax
- coding properties
- representations
- natural language understanding
- Datorlingvistik
- Computational Linguistics
Publikations- och innehållstyp
- ref (ämneskategori)
- art (ämneskategori)
Hitta via bibliotek
Till lärosätets databas