Structured, systematic-review-style extraction across many papers
Same slot as Consensus, different strength: extracting structured fields across a larger set of papers.
elicit.com — a literature-review assistant built for pulling structured fields (methods, sample, findings) out of many papers at once, closer to a lightweight systematic review than a one-off synthesis.
The same D1·3 slot as Consensus — plus D1·4/D1·5, since literature-scout accepts an Elicit extraction the same way it accepts a Semantic Scholar result, and D2·5, since a citation backed this way can end up in the manuscript.
elicit.com, sign up with email. Free tier covers the course; Plus is ≈ €13/mo.
Same test as Consensus: ask it the underlying question, then compare what it returns to the actual paper's abstract.
Consensus gives a quick synthesized answer to one question; Elicit is built for comparing methods, samples, and findings across a larger set of papers at once.
When your sub-question needs comparing several papers' methods or samples side by side, rather than one quick synthesized answer.
An extracted field is still an AI claim about a paper's content until you've opened the paper and checked — Elicit doesn't get an exemption from literature-scout's rule.
Elicit tells you what a paper reports; Scite tells you whether later work still agrees with it. Different questions, both needed.
Same Day-1 afternoon block as Consensus — finding literature gaps on your own angle, not just the shared case.
literature-scout accepts an Elicit extraction you've pasted into the chat as valid support for a citation — the same standing as a Semantic Scholar search result. Anything Elicit surfaces can end up cited in the manuscript (D2·5), and gets cross-checked again by the referee skill's citation check (D1·4).
Nothing from Elicit gets cited into a manuscript without a real, checkable source — the same literature-scout rule as everything else in this stack.
Back to D1·3's decision graph on the home page.