Үздік Бірінші Іздеу
Related Papers
TINA: A Probabilistic Syntactic Parser for Speech Understanding Systems
A new natural language system, TINA, has been developed for applications involving speech understanding tasks, which integrates key ideas from context free grammars, Augmented Transition Networks (ATN's) [1], and Lexical …...
Achieving Model Robustness through Discrete Adversarial Training
AbstractDiscrete adversarial attacks are symbolic perturbations to a language input that preserve the output label but lead to a prediction error. While such attacks have been extensively explored for the …...
Massive-scale Decoding for Text Generation using Lattices
AbstractConditional neural text generation models generate high-quality outputs, but often concentrate around a mode when what we really want is a diverse set of options. We present a search algorithm …...
Best-k Search Algorithm for Neural Text Generation
AbstractModern natural language generation paradigms require a decoding strategy to obtain quality sequences out of the model. Beam search yields high-quality but low diversity outputs; stochastic approaches suffer from high …...