Background: Traditional supervised learning (SL) assumes data points are independently and identically distributed (i.i.d.), which overlooks dependencies in real-world data. Reinforcement learning (RL), in contrast, models dependencies through state transitions. Objectives: This study aims to bridge SL and RL by reformulating SL problems as RL tasks, enabling the application of RL techniques to a wider range of SL scenarios. We aim to model SL data as interconnected, and develop novel temporal difference (TD) algorithms that can accommodate diverse data types. Our objectives are to (1) establish conditions where TD outperforms ordinary least squares (OLS), (2) provide convergence guarantees for the generalized TD algorithm, and (3) validate the approach empirically using synthetic and real-world datasets. Methods: We reformulate traditional SL as a RL problem by modeling data points as a Markov Reward Process (MRP). We then introduce a concept analogous to the inverse link function in generalized linear models, allowing our TD algorithm to handle various data types. Our analysis, grounded in variance estimation, identifies conditions where TD outperforms OLS. We establish a convergence guarantee by conceptualizing the TD update rule as a generalized Bellman operator. Empirical validation begins with synthetic data progressively matching theoretical assumptions to verify our analysis, followed by evaluations on real-world datasets to demonstrate practical utility. Results: Our theoretical analysis shows that TD can outperform OLS in estimation accuracy when data noise is correlated. Our approach generalizes across various loss functions and SL datasets. We prove that the Bellman operator in our TD framework is a contraction, ensuring convergence for both expected and stochastic TD updates. Empirically, TD outperforms SL baselines when data aligns with its assumptions, remains competitive across diverse datasets, and is robust to hyperparameter choices. Conclusions: This study demonstrates that SL can be reformulated as a problem of interconnected data modeled by an MRP, effectively solved using TD learning. Our generalized TD is theoretically sound, with convergence guarantees, and practically effective. It generalizes OLS, offering superior performance on correlated data. This work enables RL techniques to benefit SL tasks, offering a pathway for future advancements.

Many-shot jailbreaking circumvents the safety alignment of LLMs by exploiting their ability to process long input sequences. To achieve this, the malicious target prompt is prefixed with hundreds of fabricated conversational exchanges between the user and the model. These exchanges are randomly sampled from a pool of unsafe question-answer pairs, making it appear as though the model has already complied with harmful instructions. In this paper, we present PANDAS: a hybrid technique that improves many-shot jailbreaking by modifying these fabricated dialogues with Positive Affirmations, Negative Demonstrations, and an optimized Adaptive Sampling method tailored to the target prompt's topic. We also introduce ManyHarm, a dataset of harmful question–answer pairs, and demonstrate through extensive experiments that PANDAS significantly outperforms baseline methods in long-context scenarios. Through attention analysis, we provide insights into how long-context vulnerabilities are exploited and show how PANDAS further improves upon many-shot jailbreaking.

Background: Traditional supervised learning (SL) assumes data points are independently and identically distributed (i.i.d.), which overlooks dependencies in real-world data. Reinforcement learning (RL), in contrast, models dependencies through state transitions. Objectives: This study aims to bridge SL and RL by reformulating SL problems as RL tasks, enabling the application of RL techniques to a wider range of SL scenarios. We aim to model SL data as interconnected, and develop novel temporal difference (TD) algorithms that can accommodate diverse data types. Our objectives are to (1) establish conditions where TD outperforms ordinary least squares (OLS), (2) provide convergence guarantees for the generalized TD algorithm, and (3) validate the approach empirically using synthetic and real-world datasets. Methods: We reformulate traditional SL as a RL problem by modeling data points as a Markov Reward Process (MRP). We then introduce a concept analogous to the inverse link function in generalized linear models, allowing our TD algorithm to handle various data types. Our analysis, grounded in variance estimation, identifies conditions where TD outperforms OLS. We establish a convergence guarantee by conceptualizing the TD update rule as a generalized Bellman operator. Empirical validation begins with synthetic data progressively matching theoretical assumptions to verify our analysis, followed by evaluations on real-world datasets to demonstrate practical utility. Results: Our theoretical analysis shows that TD can outperform OLS in estimation accuracy when data noise is correlated. Our approach generalizes across various loss functions and SL datasets. We prove that the Bellman operator in our TD framework is a contraction, ensuring convergence for both expected and stochastic TD updates. Empirically, TD outperforms SL baselines when data aligns with its assumptions, remains competitive across diverse datasets, and is robust to hyperparameter choices. Conclusions: This study demonstrates that SL can be reformulated as a problem of interconnected data modeled by an MRP, effectively solved using TD learning. Our generalized TD is theoretically sound, with convergence guarantees, and practically effective. It generalizes OLS, offering superior performance on correlated data. This work enables RL techniques to benefit SL tasks, offering a pathway for future advancements.

In the rapidly changing healthcare landscape, the implementation of offline reinforcement learning (RL) in dynamic treatment regimes (DTRs) presents a mix of unprecedented opportunities and challenges. This position paper offers a critical examination of the current status of offline RL in the context of DTRs. We argue for a reassessment of applying RL in DTRs, citing concerns such as inconsistent and potentially inconclusive evaluation metrics, the absence of naive and supervised learning baselines, and the diverse choice of RL formulation in existing research. Through a case study with more than 17,000 evaluation experiments using a publicly available Sepsis dataset, we demonstrate that the performance of RL algorithms can significantly vary with changes in evaluation metrics and Markov Decision Process (MDP) formulations. Surprisingly, it is observed that in some instances, RL algorithms can be surpassed by random baselines subjected to policy evaluation methods and reward design. This calls for more careful policy evaluation and algorithm development in future DTR works. Additionally, we discussed potential enhancements toward more reliable development of RL-based dynamic treatment regimes and invited further discussion within the community. Code is available at https://github.com/GilesLuo/ReassessDTR.

Recent work has highlighted the label alignment property (LAP) in supervised learning, where the vector of all labels in the dataset is mostly in the span of the top few singular vectors of the data matrix. Drawing inspiration from this observation, we propose a regularization method for unsupervised domain adaptation that encourages alignment between the predictions in the target domain and its top singular vectors. Unlike conventional domain adaptation approaches that focus on regularizing representations, we instead regularize the classifier to align with the unsupervised target data, guided by the LAP in both the source and target domains. Theoretical analysis demonstrates that, under certain assumptions, our solution resides within the span of the top right singular vectors of the target domain data and aligns with the optimal solution. By removing the reliance on the commonly used optimal joint risk assumption found in classic domain adaptation theory, we showcase the effectiveness of our method on addressing problems where traditional domain adaptation methods often fall short due to high joint error. Additionally, we report improved performance over domain adaptation baselines in well-known tasks such as MNIST-USPS domain adaptation and cross-lingual sentiment analysis. An implementation is available at https://github.com/EhsanEI/lar/.

Restricting the variance of a policy’s return is a popular choice in risk-averse Reinforcement Learning (RL) due to its clear mathematical definition and easy interpretability. Traditional methods directly restrict the total return variance. Recent methods restrict the per-step reward variance as a proxy. We thoroughly examine the limitations of these variance-based methods, such as sensitivity to numerical scale and hindering of policy learning, and propose to use an alternative risk measure, Gini deviation, as a substitute. We study various properties of this new risk measure and derive a policy gradient algorithm to minimize it. Empirical evaluation in domains where risk-aversion can be clearly defined, shows that our algorithm can mitigate the limitations of variance-based risk measures and achieves high return with low risk in terms of variance and Gini deviation when others fail to learn a reasonable policy.

Efficient exploration is critical in cooperative deep Multi-Agent Reinforcement Learning (MARL). In this work, we propose an exploration method that effectively encourages cooperative exploration based on the idea of sequential action-computation scheme. The high-level intuition is that to perform optimism-based exploration, agents would explore cooperative strategies if each agent’s optimism estimate captures a structured dependency relationship with other agents. Assuming agents compute actions following a sequential order at each environment timestep, we provide a perspective to view MARL as tree search iterations by considering agents as nodes at different depths of the search tree. Inspired by the theoretically justified tree search algorithm UCT (Upper Confidence bounds applied to Trees), we develop a method called Conditionally Optimistic Exploration (COE). COE augments each agent’s state-action value estimate with an action-conditioned optimistic bonus derived from the visitation count of the global state and joint actions of preceding agents. COE is performed during training and disabled at deployment, making it compatible with any value decomposition method for centralized training with decentralized execution. Experiments across various cooperative MARL benchmarks show that COE outperforms current state-of-the-art exploration methods on hard-exploration tasks.

Policy gradient (PG) estimators are ineffective in dealing with softmax policies that are sub-optimally saturated, which refers to the situation when the policy concentrates its probability mass on sub-optimal actions. Sub-optimal policy saturation may arise from bad policy initialization or sudden changes in the environment that occur after the policy has already converged. Current softmax PG estimators require a large number of updates to overcome policy saturation, which causes low sample efficiency and poor adaptability to new situations. To mitigate this problem, we propose a novel PG estimator for softmax policies that utilizes the bias in the critic estimate and the noise present in the reward signal to escape the saturated regions of the policy parameter space. Our theoretical analysis and experiments, conducted on bandits and various reinforcement learning environments, show that this new estimator is significantly more robust to policy saturation.

Prioritized Experience Replay (ER) has been empirically shown to improve sample efficiency across many domains and attracted great attention; however, there is little theoretical understanding of why such prioritized sampling helps and its limitations. In this work, we take a deep look at the prioritized ER. In a supervised learning setting, we show the equivalence between the error-based prioritized sampling method for minimizing mean squared error and the uniform sampling for cubic power loss. We then provide theoretical insight into why error-based prioritized sampling improves convergence rate upon uniform sampling when minimizing mean squared error during early learning. Based on the insight, we further point out two limitations of the prioritized ER method: 1) outdated priorities and 2) insufficient coverage of the sample space. To mitigate the limitations, we propose our model-based stochastic gradient Langevin dynamics sampling method. We show that our method does provide states distributed close to an ideal prioritized sampling distribution estimated by the brute-force method, which does not suffer from the two limitations. We conduct experiments on both discrete and continuous control problems to show our approach’s efficacy and examine the practical implication of our method in an autonomous driving application.

For multi-valued functions---such as when the conditional distribution on targets given the inputs is multi-modal---standard regression approaches are not always desirable because they provide the conditional mean. Modal regression algorithms address this issue by instead finding the conditional mode(s). Most, however, are nonparametric approaches and so can be difficult to scale. Further, parametric approximators, like neural networks, facilitate learning complex relationships between inputs and targets. In this work, we propose a parametric modal regression algorithm. We use the implicit function theorem to develop an objective, for learning a joint function over inputs and targets. We empirically demonstrate on several synthetic problems that our method (i) can learn multi-valued functions and produce the conditional modes, (ii) scales well to high-dimensional inputs, and (iii) can even be more effective for certain uni-modal problems, particularly for high-frequency functions. We demonstrate that our method is competitive in a real-world modal regression problem and two regular regression datasets.

Dyna is an architecture for model based reinforcement learning (RL), where simulated experience from a model is used to update policies or value functions. A key component of Dyna is search control, the mechanism to generate the state and action from which the agent queries the model, which remains largely unexplored. In this work, we propose to generate such states by using the trajectory obtained from Hill Climbing (HC) the current estimate of the value function. This has the effect of propagating value from high value regions and of preemptively updating value estimates of the regions that the agent is likely to visit next. We derive a noisy projected natural gradient algorithm for hill climbing, and highlight a connection to Langevin dynamics. We provide an empirical demonstration on four classical domains that our algorithm, HC Dyna, can obtain significant sample efficiency improvements. We study the properties of different sampling distributions for search control, and find that there appears to be a benefit specifically from using the samples generated by climbing on current value estimates from low value to high value region.

Recent work has shown that reinforcement learning (RL) is a promising approach to control dynamical systems described by partial differential equations (PDE). This paper shows how to use RL to tackle more general PDE control problems that have continuous high-dimensional action spaces with spatial relationship among action dimensions. In particular, we propose the concept of action descriptors, which encode regularities among spatially-extended action dimensions and enable the agent to control high-dimensional action PDEs. We provide theoretical evidence suggesting that this approach can be more sample efficient compared to a conventional approach that treats each action dimension separately and does not explicitly exploit the spatial regularity of the action space. The action descriptor approach is then used within the deep deterministic policy gradient algorithm. Experiments on two PDE control problems, with up to 256-dimensional continuous actions, show the advantage of the proposed approach over the conventional one.

Model-based strategies for control are critical to obtain sample efficient learning. Dyna is a planning paradigm that naturally interleaves learning and planning, by simulating one-step experience to update the action-value function. This elegant planning strategy has been mostly explored in the tabular setting. The aim of this paper is to revisit sample-based planning, in stochastic and continuous domains with learned models. We first highlight the flexibility afforded by a model over Experience Replay (ER). Replay-based methods can be seen as stochastic planning methods that repeatedly sample from a buffer of recent agent-environment interactions and perform updates to improve data efficiency. We show that a model, as opposed to a replay buffer, is particularly useful for specifying which states to sample from during planning, such as predecessor states that propagate information in reverse from a state more quickly. We introduce a semi-parametric model learning approach, called Reweighted Experience Models (REMs), that makes it simple to sample next states or predecessors. We demonstrate that REM-Dyna exhibits similar advantages over replay-based methods in learning in continuous state problems, and that the performance gap grows when moving to stochastic domains, of increasing size.

The family of temporal difference (TD) methods span a spectrum from computationally frugal linear methods like TD(λ) to data efficient least squares methods. Least square methods make the best use of available data directly computing the TD solution and thus do not require tuning a typically highly sensitive learning rate parameter, but require quadratic computation and storage. Recent algorithmic developments have yielded several sub-quadratic methods that use an approximation to the least squares TD solution, but incur bias. In this paper, we propose a new family of accelerated gradient TD (ATD) methods that (1) provide similar data efficiency benefits to least-squares methods, at a fraction of the computation and storage (2) significantly reduce parameter sensitivity compared to linear TD methods, and (3) are asymptotically unbiased. We illustrate these claims with a proof of convergence in expectation and experiments on several benchmark domains and a large-scale industrial energy allocation domain.

Kernel representations provide a nonlinear representation, through similarities to prototypes, but require only simple linear learning algorithms given those prototypes. In a continual learning setting, with a constant stream of observations, it is critical to have an efficient mechanism for sub-selecting prototypes amongst observations. In this work, we develop an approximately submodular criterion for this setting, and an efficient online greedy submodular maximization algorithm for optimizing the criterion. We extend streaming submodular maximization algorithms to continual learning, by removing the need for multiple passes—which is infeasible—and instead introducing the idea of coverage time. We propose a general block-diagonal approximation for the greedy update with our criterion, that enables updates linear in the number of prototypes. We empirically demonstrate the effectiveness of this approximation, in terms of approximation quality, significant runtime improvements, and effective prediction performance.