Q: Want clarity on 20 Calls , 8 Columns , data Extraction
Dear Team , Please explain on the below three feature , how to utilize them :
20 calls in 3 hours rate limit
Literature review (beta): 8 columns
Data extractions: 25 documents per extraction
Mallik_Paperguide
Edited Nov 22, 2024A: Hi,
Thanks for your question
The AI call limit is for AI Research assistant tools which includes the prompts for literature review, data extraction requests, chat with papers/pdfs, AI search and AI Summary features. AI writer does not have any such restrictions for AI generations (depending on your subscription plan). You can generate AI writing and editing suggestions in the writer tool without any AI call restrictions.
To elaborate on AI call rate limit, one AI response is considered as AI call. For example if you are ask literature review tool to find relevant papers on climate change, it will respond with a summary based on top relevant papers and gives a list of papers along with insights. It will be considered as one AI call. If you add columns to the response (to extract summary, methodology or any custom information), it will not be treated as additional call. For chat with papers/chat with pdf, when you further probe on the initial response using a follow-up question, the subsequent response will be considered as a new AI call. So if you have 20 AI calls / 3 hours limit, you would need to wait for new Ai generations if you already have used 20 calls within a 3 hour window. After the window, the rate limit resets
In literature review and Extract data features, you can extract Eight columns in tier 1 subscription and upto 50 columns in tier 2 and tier 3 subscriptions. Each column is a parameter like summary, methodology, limitations or any custom defined detail you want to extract from the papers in literature review or Extract Data feature
When you use Extract Data feature, you can select upto 25 files at a time
You can check this video to understand how the columns can be used in literature review and Extract Data features and how Extract Data feature works
https://youtu.be/w-aD5lA6z1Y?feature=shared&t=60