Every paper here is graded on novelty, rigor and depth. Those grades turn into a map of the field: which topics are heating up, which are cooling off, and which papers in each are worth your time.
Share of papers carrying each topic, month by month. Click any topic anywhere on this page to put it on the chart.
Share of new papers compared with the same stretch before it.
Well established, but taking up less of each month's output.
Each line is drawn against that topic's own average, so above the dash is a month it ran hot and below is a month it ran cold. Shape, not size. Click one to put it on the chart above.
Each bubble is a topic. Right of the dotted line is more novel than the typical topic, above it holds up better. Bubble size is how much is being published, and color is momentum.
Each topic's share of new papers in the last 90 days against the 90 days before. Above 1 means it is taking more of what gets published. Click a bar to put that topic on the chart above.
Share, not raw counts, so a busier quarter does not lift every topic at once. A dotted line marks 1×, where a topic is holding its share.
Bar length is how many papers came from each venue. Shade is the mean quality grade of those papers, darker means stronger evidence.
Published recently, graded high on novelty and rigor, and almost untouched by citations. Citations lag by years, so these are papers you can read before the field catches up.
The strongest recent papers behind each topic above.
Built from every papers, dated by the day they were published. Scores are model-assigned evidence grades, not citation counts. High novelty with low reproducibility is a real signature and is shown on purpose. The same evidence is available through the MCP server.