Zobrazeno 1 - 10
of 58
pro vyhledávání: '"Schluter, Natalie"'
Large language models are trained on massive scrapes of the web, as required by current scaling laws. Most progress is made for English, given its abundance of high-quality pretraining data. For most other languages, however, such high quality pretra
Externí odkaz:
http://arxiv.org/abs/2411.12986
We introduce GrammaMT, a grammatically-aware prompting approach for machine translation that uses Interlinear Glossed Text (IGT), a common form of linguistic description providing morphological and lexical annotations for source sentences. GrammaMT p
Externí odkaz:
http://arxiv.org/abs/2410.18702
Autor:
Mousavi, Ali, Zhan, Xin, Bai, He, Shi, Peng, Rekatsinas, Theo, Han, Benjamin, Li, Yunyao, Pound, Jeff, Susskind, Josh, Schluter, Natalie, Ilyas, Ihab, Jaitly, Navdeep
Datasets that pair Knowledge Graphs (KG) and text together (KG-T) can be used to train forward and reverse neural models that generate text from KG and vice versa. However models trained on datasets where KG and text pairs are not equivalent can suff
Externí odkaz:
http://arxiv.org/abs/2309.11669
The central bottleneck for low-resource NLP is typically regarded to be the quantity of accessible data, overlooking the contribution of data quality. This is particularly seen in the development and evaluation of low-resource systems via down sampli
Externí odkaz:
http://arxiv.org/abs/2211.07534
Autor:
Schluter, Natalie
This paper examines the assumptions of the derived equivalence between dropout noise injection and $L_2$ regularisation for logistic regression with negative log loss. We show that the approximation method is based on a divergent Taylor expansion, ma
Externí odkaz:
http://arxiv.org/abs/1905.11320
Autor:
Varab, Daniel, Schluter, Natalie
This paper describes the design and use of the graph-based parsing framework and toolkit UniParse, released as an open-source python software package. UniParse as a framework novelly streamlines research prototyping, development and evaluation of gra
Externí odkaz:
http://arxiv.org/abs/1807.04053
Autor:
Agić, Željko, Schluter, Natalie
The recent years have seen a revival of interest in textual entailment, sparked by i) the emergence of powerful deep neural network learners for natural language processing and ii) the timely development of large-scale evaluation datasets such as SNL
Externí odkaz:
http://arxiv.org/abs/1704.05347
Autor:
Schluter, Natalie
Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.
Externí odkaz:
http://hdl.handle.net/1866/16569
Autor:
Schluter, Natalie
We present a study on lookahead hierarchies for restarting automata with auxiliary symbols and small lookahead. In particular, we show that there are just two different classes of languages recognised RRWW automata, through the restriction of lookahe
Externí odkaz:
http://arxiv.org/abs/1101.1640
Autor:
Dunlaing, Colm O., Schluter, Natalie
In 2002 Jurdzinski and Lorys settled a long-standing conjecture that palindromes are not a Church-Rosser language. Their proof required a sophisticated theory about computation graphs of 2-stack automata. We present their proof in terms of 1-tape Tur
Externí odkaz:
http://arxiv.org/abs/0710.4499