"What did I save in the last year about reinforcement learning from human feedback?"
— what happens: Returns the threads, grouped by angle, with citation links back to each tweet.
Plate ∗.for researchers
xmark for researchers.
Your X bookmarks are a literature review waiting to happen.
You save threads about your field on X. xmark turns the pile into a searchable corpus — semantic search by topic, chat with citations, export to Markdown for your notes app.
Questions that work
"Summarize the disagreements in the threads I saved about consciousness."
— what happens: Claude reads the relevant bookmarks and steelmans each side with citations.
"Find the post that introduced me to [author] — I think it was a quote tweet."
— what happens: Semantic + structural search picks up quoted-tweet context that keyword misses.
A workflow that works
- 01.Sync xmark once a week.
- 02.Ask topic questions when starting a new doc or memo.
- 03.Export results as Markdown into Obsidian / Notion / Logseq.
- 04.Cite the original tweets in your prose, not the xmark URL.
ready to get started?
sign up in seconds. free to start, no credit card required.