Seq2Seq
A neural network model based on an encoder-decoder architecture that maps an input sequence to an output sequence. It is widely used in NLP tasks including machine translation, dialogue generation, and text summarization, though it is prone to generating generic responses.
Core Facts
Timeline (last 90 days)
Traditional Seq2Seq dialogue models are deterministic and tend to produce uniform, safe responses given the same input, a phenomenon known as the "generic response" problem.
The Seq2Seq architecture compresses a source-language sentence into a vector representation via an encoder, and then a decoder progressively generates the target-language word sequence.
All Facts (3)
The Seq2Seq model was first proposed by Ilya Sutskever, Oriol Vinyals, and Quoc V. Le in the 2014 paper "Sequence to Sequence Learning with Neural Networks".
75%UnverifiedThe Seq2Seq architecture compresses a source-language sentence into a vector representation via an encoder, and then a decoder progressively generates the target-language word sequence.
60%UnverifiedTraditional Seq2Seq dialogue models are deterministic and tend to produce uniform, safe responses given the same input, a phenomenon known as the "generic response" problem.
50%