About
Who makes this site, and why
About two thousand AI papers appear on arXiv every week. Nobody reads them all, and lists of new releases do not say what is moving. This site tries to answer one question: what changed this week, and what is worth opening?
Who
Emergence is published by Charles Lerminiaux, founder of Caameleon, an artificial intelligence consultancy. The site is not attached to any laboratory, any scientific publisher or any funding tied to the topics it tracks.
The stance
Show the computations rather than hide them. Every figure shown here can be explained, and the method page says how it is obtained, including when the answer is “this is an approximation”.
Cite papers rather than trends. A digest announcing a rise without giving a title to open is of no use.
Sources and reuse
Papers come from arXiv. Abstracts remain the property of their authors: the site shows only an excerpt and always links back to the original paper on arXiv.
The classification into domains, themes and topics comes from OpenAlex, whose data is released into the public domain.
The aggregates produced here, counts, emerging signals and projections, are derived from those sources. They may be reused freely, with credit to the site.
What the site does not do
No data resale, no tracking cookie. Audience measurement is limited to an anonymous page-view count.
The research assistant answers from the indexed abstracts and cites the papers it relies on. It does not replace reading the papers, and an abstract announces a contribution without demonstrating it.
Get in touch
A figure that looks wrong, a topic filed in the wrong place, a parasitic phrase at the top of the signals: this kind of feedback fixes the site faster than any automatic tuning. That is what the contact page is for.
The site is open to sponsorship. If your organisation wants to support an independent observatory of AI research and be credited for it, the form has a reason for that.