{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/dialogwae-multimodal-response-generation-with","title":"DialogWAE: Multimodal Response Generation with Conditional Wasserstein Auto-Encoder","arxiv_id":"1805.12352","date":"2018-05-31","proceeding":"ICLR 2019 5","authors":["Xiaodong Gu","Kyunghyun Cho","Jung-Woo Ha","Sunghun Kim"],"abstract":"Variational autoencoders~(VAEs) have shown a promise in data-driven\nconversation modeling. However, most VAE conversation models match the\napproximate posterior distribution over the latent variables to a simple prior\nsuch as standard normal distribution, thereby restricting the generated\nresponses to a relatively simple (e.g., unimodal) scope. In this paper, we\npropose DialogWAE, a conditional Wasserstein autoencoder~(WAE) specially\ndesigned for dialogue modeling. Unlike VAEs that impose a simple distribution\nover the latent variables, DialogWAE models the distribution of data by\ntraining a GAN within the latent variable space. Specifically, our model\nsamples from the prior and posterior distributions over the latent variables by\ntransforming context-dependent random noise using neural networks and minimizes\nthe Wasserstein distance between the two distributions. We further develop a\nGaussian mixture prior network to enrich the latent space. Experiments on two\npopular datasets show that DialogWAE outperforms the state-of-the-art\napproaches in generating more coherent, informative and diverse responses.","url_abs":"http://arxiv.org/abs/1805.12352v2","url_pdf":"http://arxiv.org/pdf/1805.12352v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"dialogwae-multimodal-response-generation-with","repo_url":"https://github.com/MohdElgaar/DialogWAE","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"dialogwae-multimodal-response-generation-with","repo_url":"https://github.com/fangleai/Implicit-LVM","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"dialogwae-multimodal-response-generation-with","repo_url":"https://github.com/guxd/DialogWAE","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"response-generation","task_name":"Response Generation"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1805.12352","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}