Zobrazeno 1 - 10
of 10
pro vyhledávání: '"DuSell, Brian"'
Autor:
Giulianelli, Mario, Malagutti, Luca, Gastaldi, Juan Luis, DuSell, Brian, Vieira, Tim, Cotterell, Ryan
Language models are widely used in computational psycholinguistics to test theories that relate the negative log probability (the surprisal) of a region of interest (a substring of characters) under a language model to its cognitive cost experienced
Externí odkaz:
http://arxiv.org/abs/2410.02691
Autor:
Gastaldi, Juan Luis, Terilla, John, Malagutti, Luca, DuSell, Brian, Vieira, Tim, Cotterell, Ryan
Tokenization - the practice of converting strings of characters over an alphabet into sequences of tokens over a vocabulary - is a critical yet under-theorized step in the NLP pipeline. Notably, it remains the only major step not fully integrated int
Externí odkaz:
http://arxiv.org/abs/2407.11606
Computational historical linguistics seeks to systematically understand processes of sound change, including during periods at which little to no formal recording of language is attested. At the same time, few computational resources exist which deep
Externí odkaz:
http://arxiv.org/abs/2404.16341
Autor:
DuSell, Brian, Chiang, David
Attention, specifically scaled dot-product attention, has proven effective for natural language, but it does not have a mechanism for handling hierarchical patterns of arbitrary nesting depth, which limits its ability to recognize certain syntactic s
Externí odkaz:
http://arxiv.org/abs/2310.01749
Autor:
DuSell, Brian
Human language is full of compositional syntactic structures, and although neural networks have contributed to groundbreaking improvements in computer systems that process language, widely-used neural network architectures still exhibit limitations i
Externí odkaz:
http://arxiv.org/abs/2304.12955
Weighted pushdown automata (WPDAs) are at the core of many natural language processing tasks, like syntax-based statistical machine translation and transition-based dependency parsing. As most existing dynamic programming algorithms are designed for
Externí odkaz:
http://arxiv.org/abs/2210.06884
Autor:
DuSell, Brian, Chiang, David
Traditional recurrent neural networks (RNNs) have a fixed, finite number of memory cells. In theory (assuming bounded range and precision), this limits their formal language recognition power to regular languages, and in practice, RNNs have been show
Externí odkaz:
http://arxiv.org/abs/2210.01343
Autor:
DuSell, Brian, Chiang, David
Learning hierarchical structures in sequential data -- from simple algorithmic patterns to natural language -- in a reliable, generalizable way remains a challenging problem for neural language models. Past work has shown that recurrent neural networ
Externí odkaz:
http://arxiv.org/abs/2109.01982
Autor:
DuSell, Brian, Chiang, David
We present a differentiable stack data structure that simultaneously and tractably encodes an exponential number of stack configurations, based on Lang's algorithm for simulating nondeterministic pushdown automata. We call the combination of this dat
Externí odkaz:
http://arxiv.org/abs/2010.04674
This paper describes the Notre Dame Natural Language Processing Group's (NDNLP) submission to the WNGT 2019 shared task (Hayashi et al., 2019). We investigated the impact of auto-sizing (Murray and Chiang, 2015; Murray et al., 2019) to the Transforme
Externí odkaz:
http://arxiv.org/abs/1910.07134