paperlantern

yes - the blog post has a figure showing all the improvements and how big they were.

also, some times the baseline agent tries the same idea but doesn't get as big a boost as the baseline + Paper Lantern agent. We studied it and found the reason was that the baseline tries changes in isolation whereas the research-backed ideas understand the interactions between parameters and suggests multiple changes at the same time - which the baseline agent never discovers.

Thanks for trying it out :)

Yes, we contextualize the search for technques to your work and then pull everything out : the idea itself, supporting results and all implementation details - so your coding agent has a trivial time implementing it

yes - definitely, i have noticed the same that customized solutions for research papers works better than LLM web search.

thanks for catching the link issue

in case you can try out our solution for code agents, i'd love to hear what you think of it...