{"id":17964,"date":"2026-04-20T06:03:17","date_gmt":"2026-04-20T10:03:17","guid":{"rendered":"http:\/\/whamit.mit.edu\/?p=17964"},"modified":"2026-04-23T14:23:19","modified_gmt":"2026-04-23T18:23:19","slug":"minicourse-paul-smolensky-microsoft-johns-hopkins-university","status":"publish","type":"post","link":"https:\/\/ling-phil.mit.edu\/whamit\/2026\/04\/20\/minicourse-paul-smolensky-microsoft-johns-hopkins-university\/","title":{"rendered":"Minicourse - Paul Smolensky (Microsoft\/Johns Hopkins University)"},"content":{"rendered":"
Speaker<\/b>: Paul Smolensky (Microsoft\/Johns Hopkins University)<\/div>\n
Title<\/b>: \u0093Generative Linguistics meets Generative AI\u0094<\/div>\n
When<\/b>: Wednesday, April 22nd, 1pm-2:30pm (Day 1) + Thursday, April 23rd, 12:30-2pm (Day 2)\u00a0\u00a0<\/div>\n
Where<\/b>: 32-D461<\/div>\n
\u00a0<\/div>\n
Abstract: <\/strong>Do the unprecedented syntactic abilities of modern generative AI systems \u0097 Large Language Models (LLMs) \u0097 falsify traditional generative principles of linguistic knowledge, given that these principles rely heavily on symbolic computation, which appears to be absent from the neural network architectures of these LLMs? In Friday\u0092s colloquium I will argue: no. The technical foundation of this argument is introduced in the tutorial lectures of Wednesday and Thursday.<\/div>\n
\u00a0<\/div>\n
Day 1: Explaining higher cognition: Symbol structures in neural patterns<\/strong><\/div>\n
This first tutorial shows how the compositional structure of symbolic representations like those employed standardly in generative linguistics can emerge from a type of compositional structure within activation patterns in neural networks: TPRs. TPRs constitute a novel type of compositional structure that is in general continuous \u0097 non-discrete \u0097 although special cases include standard discrete structures (strings, trees, propositions). Discrete TPRs can be described as standard symbol structures that can be operated upon by neural computations to compute sophisticated discrete compositional functions such as tree-adjoining and beta-reduction. Non-discrete TPRs enable new types of formal explanations for linguistic competence (and linguistic performance), discussed primarily in Thursday\u0092s lecture. The internal numerical representations learned by modern AI systems, in a wide variety of settings, can be well approximated by TPRs, enabling compositional interpretation and control of otherwise opaque systems.<\/div>\n
\u00a0<\/div>\n
Day 2: Synergies between symbolic grammatical theory and neural network computation<\/strong><\/div>\n
This second tutorial identifies strong synergies between the grammar of symbol structures and neural network computation. The compositionally-structured neural representations presented in Wednesday\u0092s tutorial, TPRs, can be described at an abstract level as symbol structures and at a lower level as numerical activation patterns. Well-formed representations can be characterized at both levels as those that maximize Harmony, giving rise to symbolic grammar formalizations based in optimization: Harmonic Grammar and Optimality Theory. TPRs enable non-discrete compositional representations in which symbols have gradient strength and can blend together within a structural position: this enables novel types of grammatical explanation exemplified here by an analysis of French liaison and of source-markedness-controlled vowel harmony. Liaison consonants\u0092 complex alternation between presence and absence is explained from their underlying strength deficiency, and marked vowels\u0092 failure to trigger harmony is explained by the relative surface weakness resulting from their markedness.<\/div>\n","protected":false},"excerpt":{"rendered":"

Speaker: Paul Smolensky (Microsoft\/Johns Hopkins University) Title: \u0093Generative Linguistics meets Generative AI\u0094 When: Wednesday, April 22nd, 1pm-2:30pm (Day 1) + Thursday, April 23rd, 12:30-2pm (Day 2)\u00a0\u00a0 Where: 32-D461 \u00a0 Abstract: Do the unprecedented syntactic abilities of modern generative AI systems \u0097 Large Language Models (LLMs) \u0097 falsify traditional generative principles of linguistic knowledge, given that […]<\/p>\n","protected":false},"author":37,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-17964","post","type-post","status-publish","format-standard","hentry","category-announcements"],"_links":{"self":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/17964","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\/37"}],"replies":[{"embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/comments?post=17964"}],"version-history":[{"count":10,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/17964\/revisions"}],"predecessor-version":[{"id":17984,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/posts\/17964\/revisions\/17984"}],"wp:attachment":[{"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/media?parent=17964"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/categories?post=17964"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ling-phil.mit.edu\/whamit\/wp-json\/wp\/v2\/tags?post=17964"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}