DE eng

Search in the Catalogues and Directories

Page: 1 2 3 4 5
Hits 1 – 20 of 82

1
Identifying the Limits of Cross-Domain Knowledge Transfer for Pretrained Models ...
Abstract: There is growing evidence that pretrained language models improve task-specific fine-tuning not just for the languages seen in pretraining, but also for new languages and even non-linguistic data. What is the nature of this surprising cross-domain transfer? We offer a partial answer via a systematic exploration of how much transfer occurs when models are denied any information about word identity via random scrambling. In four classification tasks and two sequence labeling tasks, we evaluate baseline models, LSTMs using GloVe embeddings, and BERT. We find that only BERT shows high rates of transfer into our scrambled domains, and for classification but not sequence labeling tasks. Our analyses seek to explain why transfer succeeds for some tasks but not others, to isolate the separate contributions of pretraining versus fine-tuning, and to quantify the role of word frequency. These findings help explain where and why cross-domain transfer occurs, which can guide future studies and practical fine-tuning ... : 16 pages, 5 figures, preprint ...
Keyword: Computation and Language cs.CL; FOS Computer and information sciences
URL: https://dx.doi.org/10.48550/arxiv.2104.08410
https://arxiv.org/abs/2104.08410
BASE
Hide details
2
Decrypting Cryptic Crosswords: Semantically Complex Wordplay Puzzles as a Target for NLP ...
BASE
Show details
3
Decrypting cryptic crosswords: Semantically complex wordplay puzzles as a target for NLP ...
BASE
Show details
4
Decrypting cryptic crosswords: Semantically complex wordplay puzzles as a target for NLP ...
BASE
Show details
5
Decrypting cryptic crosswords: Semantically complex wordplay puzzles as a target for NLP ...
BASE
Show details
6
Decrypting cryptic crosswords: Semantically complex wordplay puzzles as a target for NLP ...
BASE
Show details
7
Decrypting cryptic crosswords: Semantically complex wordplay puzzles as a target for NLP ...
BASE
Show details
8
Decrypting cryptic crosswords: Semantically complex wordplay puzzles as a target for NLP ...
BASE
Show details
9
Relevance-guided Supervision for OpenQA with ColBERT ...
BASE
Show details
10
Reliable Characterizations of NLP Systems as a Social Responsibility ...
BASE
Show details
11
DynaSent: A Dynamic Benchmark for Sentiment Analysis ...
BASE
Show details
12
Keynote 3 # Speaker: Chris Potts ...
BASE
Show details
13
A probabilistic pragmatics for English singular some
In: Semantics and Linguistic Theory; Proceedings of SALT 30; 22-42 ; 2163-5951 (2021)
BASE
Show details
14
Learning Compositional Negation in Populations of Roth-Erev and Neural Agents ...
BASE
Show details
15
Modeling Subjective Assessments of Guilt in Newspaper Crime Narratives ...
BASE
Show details
16
Relevance-guided Supervision for OpenQA with ColBERT ...
BASE
Show details
17
Disrupting the Dominant Discourse: Exploring the Mentoring Experiences of Latinx Community College Students
In: Education Publications (2020)
BASE
Show details
18
Neural Natural Language Inference Models Partially Embed Theories of Lexical Entailment and Negation ...
BASE
Show details
19
No Vacuous Quantification Constraints in Syntax
In: North East Linguistics Society (2020)
BASE
Show details
20
Communication-based Evaluation for Natural Language Generation
In: Proceedings of the Society for Computation in Linguistics (2020)
BASE
Show details

Page: 1 2 3 4 5

Catalogues
3
2
12
0
0
1
1
Bibliographies
19
0
1
1
0
0
0
0
6
Linked Open Data catalogues
0
Online resources
0
0
0
0
Open access documents
46
0
0
0
0
© 2013 - 2024 Lin|gu|is|tik | Imprint | Privacy Policy | Datenschutzeinstellungen ändern