DE eng

Search in the Catalogues and Directories

Page: 1...48 49 50 51 52
Hits 1.021 – 1.029 of 1.029

1021
The Low-Dimensional Linear Geometry of Contextualized Word Representations ...
Abstract: Black-box probing models can reliably extract linguistic features like tense, number, and syntactic role from pretrained word representations. However, the manner in which these features are encoded in representations remains poorly understood. We present a systematic study of the linear geometry of contextualized word representations in ELMO and BERT. We show that a variety of linguistic features (including structured dependency relationships) are encoded in low-dimensional subspaces. We then refine this geometric picture, showing that there are hierarchical relations between the subspaces encoding general linguistic categories and more specific ones, and that low-dimensional feature encodings are distributed rather than aligned to individual neurons. Finally, we demonstrate that these linear subspaces are causally related to model behavior, and can be used to perform fine-grained manipulation of BERT's output distribution. ...
Keyword: Computational Linguistics; Language Models; Machine Learning; Machine Learning and Data Mining; Natural Language Processing
URL: https://underline.io/lecture/39841-the-low-dimensional-linear-geometry-of-contextualized-word-representations
https://dx.doi.org/10.48448/hsry-qs14
BASE
Hide details
1022
Data Collection vs. Knowledge Graph Completion: What is Needed to Improve Coverage? ...
BASE
Show details
1023
Coreference-aware Surprisal Predicts Brain Response ...
BASE
Show details
1024
Linguistic Dependencies and Statistical Dependence ...
BASE
Show details
1025
Does referent predictability affect the choice of referential form? A computational approach using masked coreference resolution ...
BASE
Show details
1026
Live Session - 2E: Linguistic Theories, Cognitive Modeling and Psycholinguistics ...
BASE
Show details
1027
Coreference-aware Surprisal Predicts Brain Response ...
BASE
Show details
1028
Synthetic Textual Features for the Large-Scale Detection of Basic-level Categories in English and Mandarin ...
BASE
Show details
1029
Pragmatic competence of pre-trained language models through the lens of discourse connectives ...
BASE
Show details

Page: 1...48 49 50 51 52

Catalogues
0
0
0
0
0
0
0
Bibliographies
0
0
0
0
0
0
0
0
0
Linked Open Data catalogues
0
Online resources
0
0
0
0
Open access documents
1.029
0
0
0
0
© 2013 - 2024 Lin|gu|is|tik | Imprint | Privacy Policy | Datenschutzeinstellungen ändern