General Strategy
Research on general strategies focuses on developing and optimizing methods for achieving specific goals across diverse domains, from mitigating online trolling to enhancing the efficiency of AI systems. Current efforts concentrate on leveraging human preferences to guide strategy selection, integrating sustainability considerations into AI development, and employing techniques like imitation learning and Bayesian optimization to improve model performance and efficiency. These advancements have significant implications for various fields, improving online community management, promoting responsible AI development, and accelerating scientific discovery through more efficient computational tools.
Papers
No-Regret Online Prediction with Strategic Experts
Omid Sadeghi, Maryam Fazel
Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback
Katherine Tian, Eric Mitchell, Allan Zhou, Archit Sharma, Rafael Rafailov, Huaxiu Yao, Chelsea Finn, Christopher D. Manning