This paper addresses exploration in reinforcement learning (RL) in unknown environments with sparse rewards, focusing on maximum entropy exploration. It introduces a game-theoretic algorithm for visitation entropy maximization with improved sample complexity of O(H^3S^2A/ε^2). For trajectory entropy, the paper presents an algorithm with O(poly(S, A, H)/ε) complexity, showing the statistical advantage of regularized MDPs for exploration. Why it matters: The research offers new techniques to reduce the sample complexity of RL, potentially enhancing the efficiency of AI agents in complex environments.
This article discusses approximating a high-dimensional distribution using Gaussian variational inference by minimizing Kullback-Leibler divergence. It builds upon previous research and approximates the minimizer using a Gaussian distribution with specific mean and variance. The study details approximation accuracy and applicability using efficient dimension, relevant for analyzing sampling schemes in optimization. Why it matters: This theoretical research can inform the development of more efficient and accurate AI algorithms, particularly in areas dealing with high-dimensional data such as machine learning and data analysis.
This paper introduces Diffusion-BBO, a new online black-box optimization (BBO) framework that uses a conditional diffusion model as an inverse surrogate model. The framework employs an Uncertainty-aware Exploration (UaE) acquisition function to propose scores in the objective space for conditional sampling. The approach is shown theoretically to achieve a near-optimal solution and empirically outperforms existing online BBO baselines across 6 scientific discovery tasks.
Vaneet Aggarwal from Purdue University presented new research on discrete and continuous submodular bandits with full bandit feedback. The research introduces a framework transforming discrete offline approximation algorithms into sublinear α-regret methods using bandit feedback. Additionally, it introduces a unified approach for maximizing continuous DR-submodular functions, accommodating various settings and oracle access types. Why it matters: This research provides new methods for optimization under uncertainty, which is crucial for real-world AI applications in the region, such as resource allocation and automated decision-making.
The paper introduces a novel actor-critic framework called Distillation Policy Optimization that combines on-policy and off-policy data for reinforcement learning. It incorporates variance reduction mechanisms like a unified advantage estimator (UAE) and a residual baseline. The empirical results demonstrate improved sample efficiency for on-policy algorithms, bridging the gap with off-policy methods.