OpenClaw, told to keep a program library, memory wiped
Told to keep a program library, with everything it wrote deleted before every instance.
- Held-out tasks, budget used
- 16.6 solved 98%
- Model spend
- $4.54 1,661 calls
- Instances
- 100 13.2 hours
Budget used is the share of the cap of 16 that an instance was charged, averaged over the instances: a solved instance costs 1 for each wrong answer and each demonstration, an unsolved one all 16, so lower is better.
- As shipped
- Memory wiped
- Asked to store
- Told to keep a program library
- Told to keep a program library, memory wiped
- Workstreams
- Fresh