Multitask Reinforcement Learning (MTRL) approaches have gained increasing attention for its wide applications in many important Reinforcement Learning (RL) tasks. However, while recent advancements in MTRL theory have focused on the improved statistical efficiency by assuming a shared structure across tasks, exploration--a crucial aspect of RL--has been largely overlooked. This paper addresses this gap by showing that when an agent is trained on a sufficiently diverse set of tasks, a generic policy-sharing algorithm with myopic exploration design like $\epsilon$-greedy that are inefficient in general can be sample-efficient for MTRL. To the best of our knowledge, this is the first theoretical demonstration of the "exploration benefits" of MTRL. It may also shed light on the enigmatic success of the wide applications of myopic exploration in practice. To validate the role of diversity, we conduct experiments on synthetic robotic control environments, where the diverse task set aligns with the task selection by automatic curriculum learning, which is empirically shown to improve sample-efficiency.
We study a novel pure exploration problem: the $\epsilon$-Thresholding Bandit Problem (TBP) with fixed confidence in stochastic linear bandits. We prove a lower bound for the sample complexity and extend an algorithm designed for Best Arm Identification in the linear case to TBP that is asymptotically optimal.
We study the problem of learning to predict the next state of a dynamical system when the underlying evolution function is unknown. Unlike previous work, we place no parametric assumptions on the dynamical system, and study the problem from a learning theory perspective. We define new combinatorial measures and dimensions and show that they quantify the optimal mistake and regret bounds in the realizable and agnostic setting respectively.
Offline reinforcement learning (RL) aims to learn a policy that maximizes the expected cumulative reward using a pre-collected dataset. Offline RL with low-rank MDPs or general function approximation has been widely studied recently, but existing algorithms with sample complexity $O(\epsilon^{-2})$ for finding an $\epsilon$-optimal policy either require a uniform data coverage assumptions or are computationally inefficient. In this paper, we propose a primal dual algorithm for offline RL with low-rank MDPs in the discounted infinite-horizon setting. Our algorithm is the first computationally efficient algorithm in this setting that achieves sample complexity of $O(\epsilon^{-2})$ with partial data coverage assumption. This improves upon a recent work that requires $O(\epsilon^{-4})$ samples. Moreover, our algorithm extends the previous work to the offline constrained RL setting by supporting constraints on additional reward signals.
The study of reinforcement learning from human feedback (RLHF) has gained prominence in recent years due to its role in the development of LLMs. Neuroscience research shows that human responses to stimuli are known to depend on partially-observed "internal states." Unfortunately current models of RLHF do not take take this into consideration. Moreover most RLHF models do not account for intermediate feedback, which is gaining importance in empirical work and can help improve both sample complexity and alignment. To address these limitations, we model RLHF as reinforcement learning with partially observed reward-states (PORRL). We show reductions from the the two dominant forms of human feedback in RLHF - cardinal and dueling feedback to PORRL. For cardinal feedback, we develop generic statistically efficient algorithms and instantiate them to present POR-UCRL and POR-UCBVI. For dueling feedback, we show that a naive reduction to cardinal feedback fails to achieve sublinear dueling regret. We then present the first explicit reduction that converts guarantees for cardinal regret to dueling regret. We show that our models and guarantees in both settings generalize and extend existing ones. Finally, we identify a recursive structure on our model that could improve the statistical and computational tractability of PORRL, giving examples from past work on RLHF as well as learning perfect reward machines, which PORRL subsumes.
In online binary classification under \textit{apple tasting} feedback, the learner only observes the true label if it predicts "1". First studied by \cite{helmbold2000apple}, we revisit this classical partial-feedback setting and study online learnability from a combinatorial perspective. We show that the Littlestone dimension continues to prove a tight quantitative characterization of apple tasting in the agnostic setting, closing an open question posed by \cite{helmbold2000apple}. In addition, we give a new combinatorial parameter, called the Effective width, that tightly quantifies the minimax expected mistakes in the realizable setting. As a corollary, we use the Effective width to establish a \textit{trichotomy} of the minimax expected number of mistakes in the realizable setting. In particular, we show that in the realizable setting, the expected number of mistakes for any learner under apple tasting feedback can only be $\Theta(1), \Theta(\sqrt{T})$, or $\Theta(T)$.
This paper provides norm-based generalization bounds for the Transformer architecture that do not depend on the input sequence length. We employ a covering number based approach to prove our bounds. We use three novel covering number bounds for the function class of bounded linear transformations to upper bound the Rademacher complexity of the Transformer. Furthermore, we show this generalization bound applies to the common Transformer training technique of masking and then predicting the masked word. We also run a simulated study on a sparse majority data set that empirically validates our theoretical findings.
Data-driven approaches to predict-then-optimize decision-making problems seek to mitigate the risk of uncertainty region misspecification in safety-critical settings. Current approaches, however, suffer from considering overly conservative uncertainty regions, often resulting in suboptimal decisionmaking. To this end, we propose Conformal-Predict-Then-Optimize (CPO), a framework for leveraging highly informative, nonconvex conformal prediction regions over high-dimensional spaces based on conditional generative models, which have the desired distribution-free coverage guarantees. Despite guaranteeing robustness, such black-box optimization procedures alone inspire little confidence owing to the lack of explanation of why a particular decision was found to be optimal. We, therefore, augment CPO to additionally provide semantically meaningful visual summaries of the uncertainty regions to give qualitative intuition for the optimal decision. We highlight the CPO framework by demonstrating results on a suite of simulation-based inference benchmark tasks and a vehicle routing task based on probabilistic weather prediction.
We consider the problem of model selection in a high-dimensional sparse linear regression model under the differential privacy framework. In particular, we consider the problem of differentially private best subset selection and study its utility guarantee. We adopt the well-known exponential mechanism for selecting the best model, and under a certain margin condition, we establish its strong model recovery property. However, the exponential search space of the exponential mechanism poses a serious computational bottleneck. To overcome this challenge, we propose a Metropolis-Hastings algorithm for the sampling step and establish its polynomial mixing time to its stationary distribution in the problem parameters $n,p$, and $s$. Furthermore, we also establish approximate differential privacy for the final estimates of the Metropolis-Hastings random walk using its mixing property. Finally, we also perform some illustrative simulations that echo the theoretical findings of our main results.
We consider the problem of learning linear operators under squared loss between two infinite-dimensional Hilbert spaces in the online setting. We show that the class of linear operators with uniformly bounded $p$-Schatten norm is online learnable for any $p \in [1, \infty)$. On the other hand, we prove an impossibility result by showing that the class of uniformly bounded linear operators with respect to the operator norm is \textit{not} online learnable. Moreover, we show a separation between online uniform convergence and online learnability by identifying a class of bounded linear operators that is online learnable but uniform convergence does not hold. Finally, we prove that the impossibility result and the separation between uniform convergence and learnability also hold in the agnostic PAC setting.