Seed it with one paper — yours, a Scholar profile, a PDF, or pasted text. Paper Monitor reads everything published overnight and hands you the few worth your time, each with a one-line why it matters, written for your research.
Free while in beta. Your first feed arrives tomorrow morning.
Aggregated token entropy beats sentence-level scores at flagging fabricated citations in long answers.
Preference tuning systematically inflates a model's expressed confidence, and the effect compounds with scale.
How it works
One paper is enough — yours, a Google Scholar profile, a PDF, or a paragraph of pasted text. No keyword lists to curate, no alerts to configure.
Every morning it sweeps OpenAlex and the preprint servers through six discovery channels, then matches semantically — meaning, not keywords.
An AI judge reads everything that matched and picks the must-reads. If nothing clears the bar, your feed says so. Zero must-reads is a valid day — we never pad the feed.
Discovery
Some papers you read because they are exactly your niche. Others because everyone in your field will be talking about them — and missing those is worse. Paper Monitor watches both axes, from the same single seed.
You shouldn't learn about a Nature paper in your field from Twitter.
Inferred from where your literature actually publishes — not a hand-picked list you have to maintain. Anything they print in your area, you see.
Anything in the big multidisciplinary venues that touches your area — before a colleague forwards it to you with "have you seen this?"
Identified from corresponding authors in your seed's citation neighborhood — so a powerhouse group's new preprint surfaces even when it drifts from your keywords.
Zero configuration. From one paper, it learns the map of your field — and keeps it current as the field moves.
No padding, no fake relevance. If nothing published overnight matters to you, the feed says exactly that — and you get your morning back. An empty feed is information too.
A separate lane for adjacent-field papers that share your methods, not your vocabulary — forecast verification from a meteorology journal, mapped onto your calibration problem. Papers you would never think to search for.
One click to Zotero, one click to the open-access PDF. Triage in the feed; read where you already read. No in-app reader to learn, no library to migrate.
Seed it with one paper — yours works best — and your first feed arrives tomorrow. Free while in beta. Questions? Write to [email protected].