Epicure: Multidimensional Flavor Structure in Food Ingredient Embeddings LLM-Augmented Data Curation Reveals Culturally and Perceptually Grounded Dimensions in Food Embeddings

Jakub Radzikowski, Josef Chen · 2026

Read Epicure: Multidimensional Flavor Structure in Food Ingredient Embeddings LLM-Augmented Data Curation Reveals Culturally and Perceptually Grounded Dimensions in Food Embeddings by Jakub Radzikowski, Josef Chen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-22776.txt.

Description

Radzikowski and Chen show that a chef's sense of flavor, texture, and cultural fit, hard to put into words, already lives inside ingredient vectors learned from recipes and food chemistry. With language-model help, they clean 6,653 raw FlavorGraph entries down to 1,032 standard ingredients. From the cleaned set, they recover at least fifteen dimensions covering taste, texture, geography, processing, and culture. The result turns a scattered cook's intuition into a measurable structure for studying how ingredients relate.

Catalog record

CREATORJakub Radzikowski, Josef Chen
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE102873
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.22776

Browse the full library · About the Tower of Babel