A resource rational agent

Inspiration

resource rational analysis

What it does

Try to reach the target with reasonable resource(token), with a clear feedback and solid self correction.

How we built it

Codex, Gpt 5.6, my brain(to study resource rational analysis model)

Challenges we ran into

Numerous issues in the process of making abstract models practical

Accomplishments that we're proud of

Make it a way to control agent's token use, plus make the target more reachable.

What we learned

there's still a distance between theory and application

What's next for A resource rational agent

Count the token it saves, improve functionality of abstracting targets.

Built With

• codex
• database
• gpt
• python