Does Claude Keep Its Own Promises?
Detailed guidelines are often used to dictate how models should behave. However, are they really "trustworthy"?
Notes on machine learning, air traffic management, and whatever else I end up thinking about.
Detailed guidelines are often used to dictate how models should behave. However, are they really "trustworthy"?
Assessing the reliability of the agentic plane spotting workflow introduced in the previous post.
Can agentic workflows help with plane spotting?
No regret learning tells us something useful about decision making when you don't know what comes next.