Gaussian Process Bandits for AI System Optimization with Selectively Verified Feedback
cs.LG, cs.AI, stat.ML
Submitted: 2026-10-13
Updated: 2026-10-13
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- Uncertainty about Uncertainty: Optimal Adaptive Algorithms for Estimating Mixtures of Unknown Coins
- Graph Neural Thompson Sampling
- Best Arm Identification with LLM Judges and Limited Human
- The Surrogate Is Not the Reward: Post-Surrogate Primary-Outcome Acquisition in Contextual Bandits
- Correlation-Aware Contextual Bandits with Surrogate Rewards for LLM Routing
- Multi-Armed Bandits With Machine Learning-Generated Surrogate Rewards
- Optimal Design for Active Preference Learning with Biased LLM Judges
- Gaussian Process Bandit Optimization with Machine Learning Predictions and Application to Hypothesis Generation
- How Many Humans Are 32 LLM Judges Worth?
- Pure Exploration Beyond Reward Feedback: The Role of Post-Action Context
- Online Learning with LLM Experts from Limited Feedback
- Best Arm Identification in Generalized Linear Bandits via Hybrid Feedback
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks