Scite — AI engine for scientific litterature

Scite-P&PS partnership

Scite and P&PS — ULB have partnered to offer all researchers (doctoral candidates, post-doctoral fellows) and academics at the Université Libre de Bruxelles a fully free 6-week trial.

outcome of the trial

• A stance from P&PS for ULB endorsing or not an extended-term license for AI tool in research;
• Topics for future training in AI for scientific litterature;
• To provide guidelines for AI-powered tools in scientific production at ULB that enhance research efficiency and comply with ethical standards.

timing

1. A technical demo for researchers is given by the creators of Scite for ULB This was an online event : check the video below
2. Start of trial period follows the technical demo — thus starts on the 13th To access the trial, registration below (same link).

how to start the trial?

1. Go to https://scite.ai/ 2. Top right "Sign up" 3. create an account with your @ulb.be address 4. keep your password safe

video technical demo

click here to access it

See how other universities are using Scite:

• University of Zurich
• Manchester University
• University of Arizona
• University of Clemson

what is scite?

Scite uses a large language model (LLM) to perform tasks related to litterature exploration, review and writing.

Among other things scite can :

• look for answers to research question in the body of papers "1.2b citation statements extracted and analyzed 187m articles, book chapters, preprints, and datasets"
• this includes papers behind a paid wall (not only open source journals)
• can handle research questions in natural language ("what is the effect of X on Y?")
• provides the answers to questions via excerpts of papers : answer are always in context.
• can write litterature reviews with citation (scite assistant)

Futher read: Josh M. Nicholson, Milo Mordaunt, Patrice Lopez, Ashish Uppala, Domenic Rosati, Neves P. Rodrigues, Peter Grabitz, Sean C. Rife; scite: A smart citation index that displays the context of citations and classifies their intent using deep learning. Quantitative Science Studies 2021; 2 (3): 882–898. doi: https://doi.org/10.1162/qss_a_00146

Coverage

Check out the database here : Scite Data in a Nushell 🡕

Screenshots

---

---

---

Some of the collaborating scientific institutions

adapted from : https://scite.ai/partners/students & https://scite.ai/news-and-press