Browse Teach Curate

Pre-trained Foundation Models for Scalable Data Analysis

How pre-trained foundation models (LLMs and multimodal PFMs) are applied across the data-analysis pipeline — structured data understanding, semantic querying, feature representation, and analytical reasoning — and the architectural patterns (prompting, fine-tuning, embedding-based retrieval) that connect models to data systems.

Dalek Text

Noname manuscript No. (will be inserted by the editor) Revisiting Data Analysis with Pre-trained

Data analysis focuses on harnessing advanced statistics, programming, and machine learning techniques to extract valuable insights from vast datasets. An increasing volume and variety of research eme…