Matches in SemOpenAlex for { <https://semopenalex.org/work/W2786039921> ?p ?o ?g. }
- W2786039921 abstract "Sequence generative adversarial networks (SeqGAN) have been used to improve conditional sequence generation tasks, for example, chit-chat dialogue generation. To stabilize the training of SeqGAN, Monte Carlo tree search (MCTS) or reward at every generation step (REGS) is used to evaluate the goodness of a generated subsequence. MCTS is computationally intensive, but the performance of REGS is worse than MCTS. In this paper, we propose stepwise GAN (StepGAN), in which the discriminator is modified to automatically assign scores quantifying the goodness of each subsequence at every generation step. StepGAN has significantly less computational costs than MCTS. We demonstrate that StepGAN outperforms previous GAN-based methods on both synthetic experiment and chit-chat dialogue generation." @default.
- W2786039921 created "2018-02-23" @default.
- W2786039921 creator A5016343106 @default.
- W2786039921 creator A5040508737 @default.
- W2786039921 date "2018-08-16" @default.
- W2786039921 modified "2023-09-24" @default.
- W2786039921 title "Improving Conditional Sequence Generative Adversarial Networks by Stepwise Evaluation" @default.
- W2786039921 cites W1591706642 @default.
- W2786039921 cites W1757796397 @default.
- W2786039921 cites W2101105183 @default.
- W2786039921 cites W2119717200 @default.
- W2786039921 cites W2121863487 @default.
- W2786039921 cites W2145339207 @default.
- W2786039921 cites W2328886022 @default.
- W2786039921 cites W2410983263 @default.
- W2786039921 cites W2487501366 @default.
- W2786039921 cites W2565378226 @default.
- W2786039921 cites W2593383075 @default.
- W2786039921 cites W2605135824 @default.
- W2786039921 cites W2613904329 @default.
- W2786039921 cites W2620623908 @default.
- W2786039921 cites W2622385665 @default.
- W2786039921 cites W2626778328 @default.
- W2786039921 cites W2795601629 @default.
- W2786039921 cites W2950178297 @default.
- W2786039921 cites W2950304420 @default.
- W2786039921 cites W2950672580 @default.
- W2786039921 cites W2951520714 @default.
- W2786039921 cites W2951824008 @default.
- W2786039921 cites W2951990924 @default.
- W2786039921 cites W2962938178 @default.
- W2786039921 cites W2963248296 @default.
- W2786039921 cites W2963521413 @default.
- W2786039921 cites W2964017345 @default.
- W2786039921 cites W2964201867 @default.
- W2786039921 cites W2964268978 @default.
- W2786039921 cites W635530177 @default.
- W2786039921 doi "https://doi.org/10.48550/arxiv.1808.05599" @default.
- W2786039921 hasPublicationYear "2018" @default.
- W2786039921 type Work @default.
- W2786039921 sameAs 2786039921 @default.
- W2786039921 citedByCount "2" @default.
- W2786039921 countsByYear W27860399212019 @default.
- W2786039921 countsByYear W27860399212020 @default.
- W2786039921 crossrefType "posted-content" @default.
- W2786039921 hasAuthorship W2786039921A5016343106 @default.
- W2786039921 hasAuthorship W2786039921A5040508737 @default.
- W2786039921 hasBestOaLocation W27860399211 @default.
- W2786039921 hasConcept C105795698 @default.
- W2786039921 hasConcept C113174947 @default.
- W2786039921 hasConcept C119857082 @default.
- W2786039921 hasConcept C134306372 @default.
- W2786039921 hasConcept C137877099 @default.
- W2786039921 hasConcept C154945302 @default.
- W2786039921 hasConcept C19499675 @default.
- W2786039921 hasConcept C2778112365 @default.
- W2786039921 hasConcept C2779803651 @default.
- W2786039921 hasConcept C33923547 @default.
- W2786039921 hasConcept C34388435 @default.
- W2786039921 hasConcept C37736160 @default.
- W2786039921 hasConcept C39890363 @default.
- W2786039921 hasConcept C41008148 @default.
- W2786039921 hasConcept C46149586 @default.
- W2786039921 hasConcept C54355233 @default.
- W2786039921 hasConcept C76155785 @default.
- W2786039921 hasConcept C86803240 @default.
- W2786039921 hasConcept C94915269 @default.
- W2786039921 hasConceptScore W2786039921C105795698 @default.
- W2786039921 hasConceptScore W2786039921C113174947 @default.
- W2786039921 hasConceptScore W2786039921C119857082 @default.
- W2786039921 hasConceptScore W2786039921C134306372 @default.
- W2786039921 hasConceptScore W2786039921C137877099 @default.
- W2786039921 hasConceptScore W2786039921C154945302 @default.
- W2786039921 hasConceptScore W2786039921C19499675 @default.
- W2786039921 hasConceptScore W2786039921C2778112365 @default.
- W2786039921 hasConceptScore W2786039921C2779803651 @default.
- W2786039921 hasConceptScore W2786039921C33923547 @default.
- W2786039921 hasConceptScore W2786039921C34388435 @default.
- W2786039921 hasConceptScore W2786039921C37736160 @default.
- W2786039921 hasConceptScore W2786039921C39890363 @default.
- W2786039921 hasConceptScore W2786039921C41008148 @default.
- W2786039921 hasConceptScore W2786039921C46149586 @default.
- W2786039921 hasConceptScore W2786039921C54355233 @default.
- W2786039921 hasConceptScore W2786039921C76155785 @default.
- W2786039921 hasConceptScore W2786039921C86803240 @default.
- W2786039921 hasConceptScore W2786039921C94915269 @default.
- W2786039921 hasLocation W27860399211 @default.
- W2786039921 hasOpenAccess W2786039921 @default.
- W2786039921 hasPrimaryLocation W27860399211 @default.
- W2786039921 hasRelatedWork W1710476689 @default.
- W2786039921 hasRelatedWork W2888506554 @default.
- W2786039921 hasRelatedWork W2899838535 @default.
- W2786039921 hasRelatedWork W2939496611 @default.
- W2786039921 hasRelatedWork W2949399848 @default.
- W2786039921 hasRelatedWork W2963110372 @default.
- W2786039921 hasRelatedWork W2963744666 @default.
- W2786039921 hasRelatedWork W2964268978 @default.
- W2786039921 hasRelatedWork W2972732298 @default.
- W2786039921 hasRelatedWork W2975288815 @default.
- W2786039921 hasRelatedWork W2986864699 @default.