{"id":26029,"date":"2023-05-11T09:24:21","date_gmt":"2023-05-11T07:24:21","guid":{"rendered":"https:\/\/lium.univ-lemans.fr\/?p=26029"},"modified":"2023-05-11T11:24:35","modified_gmt":"2023-05-11T09:24:35","slug":"seminaire-de-simon-guillot-et-thibault-prouteau","status":"publish","type":"post","link":"https:\/\/lium.univ-lemans.fr\/en\/seminaire-de-simon-guillot-et-thibault-prouteau\/","title":{"rendered":"S\u00e9minaire de Simon Guillot et Thibault Prouteau"},"content":{"rendered":"

Seminare from Simon Guillot and Thibault Prouteau, PhD students at LIUM <\/h2>\n

 <\/p>\n

Date:<\/strong> 12\/05\/2023
\nTime:<\/strong> 11h00
\nLocalization:<\/strong> IC2, boardroom
\nSpeakers:<\/strong> Simon Guillot<\/a> and Thibault Prouteau<\/a>
\n 
\n <\/p>\n

Sparser is better: one step closer to word embedding interpretability<\/strong><\/p>\n

 
\n <\/p>\n

Sparse word embeddings models (SPINE, SINr) are designed to embed words in interpretable dimensions. An interpretable dimension is such that a human can interpret the semantic (or syntactic) relations between words active for a dimension. These models are useful for critical downstream tasks in natural language processing (e.g. medical or legal NLP), and digital humanities applications.<\/p>\n

The study presented in this seminar aims to extend interpretability at the vector level by integrating psycholinguistic constraints into the definition of representations. Subsequently, one of the key criteria to an interpretable model is sparsity: in order to be interpretable, not every word should be represented by all the features of the model, especially if humans have to interpret these features and their relations. This raises one question: to which extent is sparsity sustainable with regard to performance?<\/p>\n

We thus introduce a sparsification procedure to evaluate its impact on two interpretable methods (SPINE and SINr) to tend towards sustainable vector interpretability. We also introduce stability as a new criterion to interpretability.<\/p><\/div><\/div><\/div><\/div>

 <\/div>
 <\/div><\/div><\/div>","protected":false},"excerpt":{"rendered":"

Seminare from Simon Guillot and Thibault Prouteau, PhD students at LIUM   Date: 12\/05\/2023 Time: 11h00 Localization: IC2, boardroom Speakers: Simon Guillot and Thibault Prouteau     Sparser is better: one step closer to word embedding interpretability     Sparse word embeddings models (SPINE, SINr) are designed to embed words in interpretable dimensions. An interpretable […]<\/p>\n

READ MORE<\/span><\/i><\/span><\/a><\/p>\n","protected":false},"author":14,"featured_media":13238,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[43],"tags":[49],"acf":[],"_links":{"self":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/posts\/26029"}],"collection":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/users\/14"}],"replies":[{"embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/comments?post=26029"}],"version-history":[{"count":0,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/posts\/26029\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/media\/13238"}],"wp:attachment":[{"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/media?parent=26029"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/categories?post=26029"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lium.univ-lemans.fr\/en\/wp-json\/wp\/v2\/tags?post=26029"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}