The web is most useful when a document can be read without negotiation. No account, no application shell, and no hidden request is needed to reach the text on this page.
A document first
Each note in this archive is built into a complete HTML document. Its title, summary, author, date, headings, references, and canonical address are present in the first response.
That arrangement is intentionally ordinary. It works for a person using a browser, a search engine building an index, an RSS reader following new work, and an agent looking for a source it can cite.
Structure is part of the argument
Clear structure is not decoration. A useful research note should make its central claim legible, separate evidence from inference, and let a reader find the relevant section quickly.
The same principle applies to quantitative work. A model should expose its assumptions. For example, even a familiar expression such as Bayes’ theorem carries a statement about how evidence changes belief:
An archive, not a feed
Feeds optimize for the moment of publication. An archive optimizes for retrieval later. The goal here is therefore not to publish often, but to publish notes that remain understandable, linkable, and open to revision.
That is the standard this site is designed to support.