π¦ ClawHub
Oraclaw Bandit
by @whatsonyourmind
A/B testing and feature optimization for AI agents. Pick the best option automatically using Multi-Armed Bandits and Contextual Bandits (LinUCB). No data war...
π Constraints
1. Always include historical data when available β more data = better selections
2. Use ucb1 algorithm for most cases. Use thompson when you need more exploration early on.
3. Record rewards after each decision to improve future selections
4. Context vectors must be consistent length across all calls
5. Rewards should be normalized to 0-1 range
TERMINAL
clawhub install oraclaw-bandit