This is a project trying to enable LLM to generate 100 bars multi-instrument music validated by the statistical cloud</p>\n","updatedAt":"2026-06-23T16:27:10.509Z","author":{"_id":"6573ad7706fdcd4ca914564a","avatarUrl":"/avatars/64b65a0f545833463c8b3e258496e7a5.svg","fullname":"Yichen Xu","name":"xuyichen","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8660554885864258},"editors":["xuyichen"],"editorAvatarUrls":["/avatars/64b65a0f545833463c8b3e258496e7a5.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.22708","authors":[{"_id":"6a3ab3770a86ac3098d5d488","name":"Yichen Xu","hidden":false}],"publishedAt":"2026-06-21T00:00:00.000Z","submittedOnDailyAt":"2026-06-23T00:00:00.000Z","title":"Libretto: Giving LLM Agents a Sense of Musical Structure","submittedOnDailyBy":{"_id":"6573ad7706fdcd4ca914564a","avatarUrl":"/avatars/64b65a0f545833463c8b3e258496e7a5.svg","isPro":false,"fullname":"Yichen Xu","user":"xuyichen","type":"user","name":"xuyichen"},"summary":"Generative music systems can now produce impressive audio from text prompts, but audio outputs are difficult to inspect, edit, and diagnose as musical structure. We introduce Libretto, an agent-facing framework for symbolic music generation and revision. Libretto uses an LLM-native grammar with explicit onset slots, voices, and bar-level organization, then evaluates each piece in a corpus-calibrated statistical space over rhythm, harmony, melody, texture, form, and variation. The same structural axes support retrieval, diagnosis, copy-risk control, and iterative self-revision. Across gap filling, reference-guided full-piece generation, gradual morphing, and educational music generation, Libretto turns symbolic music from a raw token sequence into a measurable and editable object for language-model agents.","upvotes":1,"discussionId":"6a3ab3770a86ac3098d5d489","projectPage":"https://libretto.site/","githubRepo":"https://github.com/Xyc-arch/Libretto","githubRepoAddedBy":"user","ai_summary":"Libretto provides a structured framework for symbolic music generation and revision using LLM-native grammar and statistical evaluation across musical dimensions.","ai_keywords":["symbolic music generation","LLM-native grammar","onset slots","voices","bar-level organization","corpus-calibrated statistical space","rhythm","harmony","melody","texture","form","variation","retrieval","copy-risk control","iterative self-revision"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","githubStars":1,"organization":{"_id":"66b1baeff10262fc4fa61961","name":"UCBerkeley","fullname":"University of California, Berkeley","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/63f425c3a096536aeab42dea/bxNKEkprdm5JI1wkjmNAL.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6573ad7706fdcd4ca914564a","avatarUrl":"/avatars/64b65a0f545833463c8b3e258496e7a5.svg","isPro":false,"fullname":"Yichen Xu","user":"xuyichen","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"66b1baeff10262fc4fa61961","name":"UCBerkeley","fullname":"University of California, Berkeley","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/63f425c3a096536aeab42dea/bxNKEkprdm5JI1wkjmNAL.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.22708.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2606.22708 in a model README.md to link it from this page.

Cite arxiv.org/abs/2606.22708 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2606.22708 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.