We added sections to Epsilon literature reviews! For any scientific topic, our generated reports now highlight key findings, methodologies, challenges, and opportunities.

Percentage of published CS papers that have 0 citations An interesting trend between 2000 and 2010

There may not be a more toxic metric in any industry than the "m-index" in academia

In climate science, even a 1% improvement in model accuracy can reduce damage during extreme events and save lives. In our latest blog post, we break down 5 examples of how ML and generative AI techniques help climate researchers save the planet 🌎 epsilon-ai.com/blog/ai-climat…

Had a great time talking about AI and scientific research on the Smart Growth Rocket podcast! Francesca and I chatted about AI's impact on education, breaking into entrepreneurship, and Epsilon's mission to summarize the world's scientific knowledge! open.spotify.com/episode/1do1eW…

Today we launched Paper Focus! 🚨 From a search result, you can now easily deep dive into specific papers. Just click the "focus" button next to the paper and all follow up searches will use information directly from that article.

Today we launched a new way to ask questions within papers 🚨 Each answer contains references to passages within the text and clicking on one highlights the corresponding passage in the article. This is great for extracting specific information from dense academic articles!

The tradeoff of allowing science to move faster through open access/pre-print publications is that it increases the amount of low quality research. I think there's a lot of room for LLMs to help with screening papers, i.e. helping you decide what's worth reading vs skipping.