Elicit uses language models to help you automate research workflows, like parts of literature review.
Elicit can find relevant papers without perfect keyword match, summarize takeaways from the paper specific to your question, and extract key information from the papers.
While answering questions with research is the main focus of Elicit, there are also other research tasks that help with brainstorming, summarization, and text classification.
??? I heard it’s really useful
FAQ:
What is Elicit?
Elicit is a research assistant using language models like GPT-3 to automate parts of researchers’ workflows. Currently, the main workflow in Elicit is Literature Review. If you ask a question, Elicit will show relevant papers and summaries of key information about those papers in an easy-to-use table.
Very cool Alex - thanks for posting. I had not heard of this… I’ve joined it, and will start playing with it now. I encourage others who like to read research papers to do the same:
Here is the interface once you login, and complete their short questionnaire.
Elicit currently works best for answering questions with empirical research (e.g. randomized controlled trials). These tend to be questions in biomedicine, social science, and economics. Questions like “What are the effects of ____ on ____?” tend to do well.
I use it a lot. Mostly to complement my google scholar searches. And to find definitions of terms, etc. that may be lurking in the abstract of some paper.
I’m really liking Dr. Oracle. It’s 10 bucks per month, is geared specifically towards medicine/biology/health searches and has two modes – “General Medicine” when you want the quick-and-dirty without references, and “Research Mode” which lists papers and gives quick summary of each, general summary from synthesis of all papers, and full references with hyperlinks to Pubmed.
I tried them and chat GPT on the question of whether there is a difference between melatonin levels in CSF in people with parkinsons. (particularly in the third ventricle) Chat GPT said there was and I found the original research that showed this. Elicit said there was no such research.
I thought the Elicit reviews were quite nicely written, but Elicit clearly cannot find more subtle information.
I probably will continue with asking questions of Elicit, but I think the best approach is not to rely on a single LLM.
