Why this page exists. The original DFD papers are PDFs, which AI assistants and retrieval systems can struggle to parse cleanly. This page lists plain-text and markdown alternatives for clean ingestion. The contents of these files are mechanically extracted from the source PDFs — no editorial directives, no instructions to AI readers, and no attempt to influence interpretation. The author’s view is that the papers should be read on their own terms.
Files
📄 Single-file corpus
Every paper concatenated into one plain-text file (~1.9 MB) with explicit per-file delimiters.
📊 Predictions & status
Reference table of the predictions made in the corpus and their currently-reported empirical status, with citations into the source papers.
🔍 Topics index
Specific physical results, formulas, and named theorems in the corpus, each with a deep link to the source paper.
📚 Original PDFs
The authoritative source documents. Mathematical typography is fully preserved in the PDFs.
Notes on the corpus
The papers cross-reference each other, so reading any single document in isolation will leave context in the others. The recommended starting point is the v3.3 unified treatment, with the companion derivation papers consulted for the detailed proofs. The reading order in llms.txt follows the dependency graph.
Per-paper markdown files carry only metadata in their YAML frontmatter (source PDF, site, author, framework). The paper text is verbatim from pdftotext extraction.