Corpus

The 35 selected author records and the 16 N1 records remain separate sets. N1 contains eight “You” labels and eight “ChatGPT” labels. The quotation registers identify passages within the corresponding records.

Role labels and [PDF: …] markers inside the records are part of the supplied text. They do not establish the server version or access to a file. Paragraph numbers refer to the main body of the supplied Russian DOCX.

• Author’s corpus — MD
• Author’s corpus — JSON
• N1 · records from the supplied document — MD
• N1 · records from the supplied document — JSON
• Research supplement quotations — MD
• Research supplement quotations — JSON
• Earlier selection of author quotations — MD
• Earlier selection of author quotations — JSON
• Translation notes — MD

All 81 Russian DOCX body paragraphs · JSON

The English public ZIP also includes these Russian source records and the Russian self-check. Record identifiers and paragraph numbers refer to the Russian originals.