> For example, one could analyze _actual_ distributions or whatever (instead of assuming uniform) Postgres keeps histograms (including N most common values) fo

> For example, one could analyze _actual_ distributions or whatever (instead of assuming uniform) Postgres keeps histograms (including N most common values) for all columns; it does not blindly assume uniform distributions. (Presumably an LLM would have access to the same histograms.)