Projects

Legommenders

2026.06.15 Lead 2024–

A content-recommendation research library for the LLM era, designed for modularity, reuse, and expressive experimentation.

Legommenders is more than a code dump. I think of it as a research expression toolkit: a way to organize content recommendation, feature modeling, and large-model interfaces into a structure that is faster to explore and easier to reproduce.

I want projects to outlive a single paper and become better starting points for the next question.