I have a question about Research Agents that may be useful for other people working with it too.
I am particularly interested in evaluation and performance. I would like to compare approaches that people have actually used rather than just theoretical suggestions.
What has worked best for you...
