{"url":"/method/mixture-of-softmaxes","slug":"mixture-of-softmaxes","name":"Mixture of Softmaxes","full_name":"Mixture of Softmaxes","full_name_withheld":false,"description_markdown":"**Mixture of Softmaxes** performs $K$ different softmaxes and mixes them. The motivation is that the traditional [softmax](https://paperswithcode.com/method/softmax) suffers from a softmax bottleneck, i.e. the expressiveness of the conditional probability we can model is constrained by the combination of a dot product and the softmax. By using a mixture of softmaxes, we can model the conditional probability more expressively.","description_state":"present","introduced_year":null,"introduced_by":{"title":null,"paper":null,"first_author":null,"n_authors":0,"url_abs":null,"archive_paper_url":null},"source":{"url":"http://arxiv.org/abs/1711.03953v4","title":"Breaking the Softmax Bottleneck: A High-Rank RNN Language Model","url_on_a_paper_host":true},"code_snippet_url":"https://github.com/zihangdai/mos/blob/6f89b283878bf17541672391d098c9e6c27b09cf/model.py#L103","code_snippet_url_on_a_code_host":true,"categories":[{"area":"General","area_id":"general","collection":"Output Functions","url":"/methods/category/output-functions","pwc_aliases":[]}],"n_papers_tagged":3,"archive_num_papers":null,"papers_newest_first":[{"paper":null,"title":"Mixtape: Breaking the Softmax Bottleneck Efficiently","date":"2019-12-01","arxiv_id":null,"n_code_links":0,"syntology":null},{"paper":"/paper/fast-and-simple-mixture-of-softmaxes-with-bpe","title":"Fast and Simple Mixture of Softmaxes with BPE and Hybrid-LightRNN for Language Generation","date":"2018-09-25","arxiv_id":"1809.09296","n_code_links":1,"syntology":null},{"paper":"/paper/breaking-the-softmax-bottleneck-a-high-rank","title":"Breaking the Softmax Bottleneck: A High-Rank RNN Language Model","date":"2017-11-10","arxiv_id":"1711.03953","n_code_links":9,"syntology":{"ran":1,"of":23,"unverified":22,"pointer_only":0}}],"papers_shown":3,"tasks":[{"task":"/task/language-modeling","name":"Language Modeling","papers":2},{"task":"/task/language-modelling","name":"Language Modelling","papers":2},{"task":"/task/machine-translation","name":"Machine Translation","papers":2},{"task":"/task/translation","name":"Translation","papers":2},{"task":"/task/image-captioning","name":"Image Captioning","papers":1},{"task":"/task/text-generation","name":"Text Generation","papers":1},{"task":"/task/tree-decomposition","name":"Tree Decomposition","papers":1},{"task":"/task/high","name":"Vocal Bursts Intensity Prediction","papers":1},{"task":"/task/word-embeddings","name":"Word Embeddings","papers":1}],"tasks_shown":9,"n_tasks":9,"usage_by_year":[{"year":"2017","papers":1},{"year":"2018","papers":1},{"year":"2019","papers":1}],"row_source":"embedded","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/mixture-of-softmaxes"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}