{"id":15562,"date":"2023-04-24T06:15:02","date_gmt":"2023-04-24T10:15:02","guid":{"rendered":"http:\/\/whamit.mit.edu\/?p=15562"},"modified":"2023-04-24T09:10:43","modified_gmt":"2023-04-24T13:10:43","slug":"exp-comp-4-28-canaan-breiss-mit","status":"publish","type":"post","link":"https:\/\/ling-phil.mit.edu\/whamit\/2023\/04\/24\/exp-comp-4-28-canaan-breiss-mit\/","title":{"rendered":"Exp\/Comp 4\/28 - Kinan Martin &#038; Canaan Breiss (MIT)"},"content":{"rendered":"<div class=\"elementToProof\">\u00a0<\/div>\n<div class=\"elementToProof\"><b>Date\/Time: <\/b>Friday 4\/28 from 2-3:30pm<\/div>\n<div class=\"elementToProof\"><b>Location:<\/b> 32-D831\u00a0<\/div>\n<div class=\"elementToProof\"><b>Speaker:<\/b> Kinan Martin &amp; Canaan Breiss (MIT)<\/div>\n<div class=\"elementToProof\"><b>Title:<\/b> Probing self-supervised speech models for phonetic and phonemic information: a case study in aspiration<\/div>\n<div class=\"elementToProof\">\u00a0<\/div>\n<div class=\"elementToProof\"><b>Abstract:<\/b> Textless self-supervised speech models have grown in capabilities in recent years, but the nature of the linguistic information they encode has not yet been thoroughly examined. We evaluate the extent to which these models\u0092 learned representations align with basic representational distinctions made by humans, focusing on a set of phonetic (low-level) and phonemic (more abstract) contrasts instantiated in word-initial stops. We find that robust representations of both phonetic and phonemic distinctions emerge in early layers of these models\u0092 architectures, and are preserved in the principal components of deeper layer representations. Our findings show that speech-trained HuBERT derives a low-noise and low-dimensional sub-space corresponding to abstract phonological distinctions.<\/div>\n<div class=\"elementToProof\">\u00a0<\/div>\n","protected":false},"excerpt":{"rendered":"<p>\u00a0 Date\/Time: Friday 4\/28 from 2-3:30pm Location: 32-D831\u00a0 Speaker: Kinan Martin &amp; Canaan Breiss (MIT) Title: Probing self-supervised speech models for phonetic and phonemic information: a case study in aspiration \u00a0 Abstract: Textless self-supervised speech models have grown in capabilities in recent years, but the nature of the linguistic information they encode has not yet [&hellip;]<\/p>\n","protected":false},"author":33,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-15562","post","type-post","status-publish","format-standard","hentry","category-announcements"],"_links":{"self":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/15562","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/users\/33"}],"replies":[{"embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/comments?post=15562"}],"version-history":[{"count":2,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/15562\/revisions"}],"predecessor-version":[{"id":15573,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/15562\/revisions\/15573"}],"wp:attachment":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/media?parent=15562"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/categories?post=15562"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/tags?post=15562"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}