Publication Date
2017
Abstract
This paper presents results from a series of simulations that attempted to teach a vanilla sequence-to-sequence neural network a reduplication process. These attempts did not succeed, suggesting that added machinery is necessary for connectionist models to perform such a task.
DOI
https://doi.org/10.7275/R5N877Z9
Included in
Computational Linguistics Commons, Morphology Commons, Psycholinguistics and Neurolinguistics Commons
Comments
Comments are welcome. Please email the author at blprickett@outlook.com.