experimental design
Experimental design in AI encompasses the planning of experiments to systematically investigate hypotheses, establish criteria for model effectiveness, and reduce bias, ensuring robust and replicable results.
- ALINE: Joint Amortization for Bayesian Inference and Active Data Acquisition
- Deep Value Benchmark: Measuring Whether Models Generalize Deep values or Shallow Preferences
- Foundation Models for Scientific Discovery: From Paradigm Enhancement to Paradigm Transition
- Gemstones: A Model Suite for Multi-Faceted Scaling Laws
- Improved Regret Bounds for Linear Bandits with Heavy-Tailed Rewards
- Online Experimental Design With Estimation-Regret Trade-off Under Network Interference
- Reverse-Annealed Sequential Monte Carlo for Efficient Bayesian Optimal Experiment Design