A: Hi Matt,

Thanks for your question. You can find our roadmap here:
https://paperguideroadmap.notion.site/

Regarding comparison with elicit, we are also soon launching Deep Research Reports (expected in 2-3 days) and you can check and compare for yourself on the quality post that. We definitely have some finetuning to be done in the current AI search and literature review features which we are working on. Regarding your specific example though, the issue seems to be due to the fact that the book is not open access and even its first page/absctract is embedded as a screenshot within the web page. In such cases, it alternatively picked up the content in the URL (https://academic.oup.com/pq/article-abstract/45/180/375/1549588?redirectedFrom=fulltext). While we are trying to minimize such errors, this is an iterative process as the LLMs are usually trained on more generic scenarios and edge cases continue to be added to the prompts from our end. Currently our system works well in case of most common publishers where documents are open access and also uses the abstract in case of papers behind paywall. But in case of some websites where security blockers are there or in cases like the above where the abstract is embedded and not directly available as text, we do face issues like this. I have added this case as a ticket with our engineering team to check the issue in detail and address it.

You can reach out to us for any further help on our support chat in the help section of the website or by writing to us at hello@paperguide.ai