Creating an effective representation space is crucial for mitigating the curse of dimensionality, enhancing model generalization, addressing data sparsity, and leveraging classical models more effectively. Recent advancements in automated feature engineering (AutoFE) have made significant progress in addressing various challenges associated with representation learning, issues such as heavy reliance on intensive labor and empirical experiences, lack of explainable explicitness, and inflexible feature space reconstruction embedded into downstream tasks. However, these approaches are constrained by: 1) generation of potentially unintelligible and illogical reconstructed feature spaces, stemming from the neglect of expert-level cognitive processes; 2) lack of systematic exploration, which subsequently results in slower model convergence for identification of optimal feature space. To address these, we introduce an interaction-aware reinforced generation perspective. We redefine feature space reconstruction as a nested process of creating meaningful features and controlling feature set size through selection. We develop a hierarchical reinforcement learning structure with cascading Markov Decision Processes to automate feature and operation selection, as well as feature crossing. By incorporating statistical measures, we reward agents based on the interaction strength between selected features, resulting in intelligent and efficient exploration of the feature space that emulates human decision-making. Extensive experiments are conducted to validate our proposed approach.

We study zero-shot generalization in reinforcement learning-optimizing a policy on a set of training tasks to perform well on a similar but unseen test task. To mitigate overfitting, previous work explored different notions of invariance to the task. However, on problems such as the ProcGen Maze, an adequate solution that is invariant to the task visualization does not exist, and therefore invariance-based approaches fail. Our insight is that learning a policy that effectively $\textit{explores}$ the domain is harder to memorize than a policy that maximizes reward for a specific task, and therefore we expect such learned behavior to generalize well; we indeed demonstrate this empirically on several domains that are difficult for invariance-based approaches. Our $\textit{Explore to Generalize}$ algorithm (ExpGen) builds on this insight: we train an additional ensemble of agents that optimize reward. At test time, either the ensemble agrees on an action, and we generalize well, or we take exploratory actions, which generalize well and drive us to a novel part of the state space, where the ensemble may potentially agree again. We show that our approach is the state-of-the-art on tasks of the ProcGen challenge that have thus far eluded effective generalization, yielding a success rate of $83\%$ on the Maze task and $74\%$ on Heist with $200$ training levels. ExpGen can also be combined with an invariance based approach to gain the best of both worlds, setting new state-of-the-art results on ProcGen.

We propose Pgx, a suite of board game reinforcement learning (RL) environments written in JAX and optimized for GPU/TPU accelerators. By leveraging JAX's auto-vectorization and parallelization over accelerators, Pgx can efficiently scale to thousands of simultaneous simulations over accelerators. In our experiments on a DGX-A100 workstation, we discovered that Pgx can simulate RL environments 10-100x faster than existing implementations available in Python. Pgx includes RL environments commonly used as benchmarks in RL research, such as backgammon, chess, shogi, and Go. Additionally, Pgx offers miniature game sets and baseline models to facilitate rapid research cycles. We demonstrate the efficient training of the Gumbel AlphaZero algorithm with Pgx environments. Overall, Pgx provides high-performance environment simulators for researchers to accelerate their RL experiments. Pgx is available at https://github.com/sotetsuk/pgx.

We explore methodologies to improve the robustness of generative adversarial imitation learning (GAIL) algorithms to observation noise. Towards this objective, we study the effect of local Lipschitzness of the discriminator and the generator on the robustness of policies learned by GAIL. In many robotics applications, the learned policies by GAIL typically suffer from a degraded performance at test time since the observations from the environment might be corrupted by noise. Hence, robustifying the learned policies against the observation noise is of critical importance. To this end, we propose a regularization method to induce local Lipschitzness in the generator and the discriminator of adversarial imitation learning methods. We show that the modified objective leads to learning significantly more robust policies. Moreover, we demonstrate -- both theoretically and experimentally -- that training a locally Lipschitz discriminator leads to a locally Lipschitz generator, thereby improving the robustness of the resultant policy. We perform extensive experiments on simulated robot locomotion environments from the MuJoCo suite that demonstrate the proposed method learns policies that significantly outperform the state-of-the-art generative adversarial imitation learning algorithm when applied to test scenarios with noise-corrupted observations.

Sampling-based Model Predictive Control (MPC) has been a practical and effective approach in many domains, notably model-based reinforcement learning, thanks to its flexibility and parallelizability. Despite its appealing empirical performance, the theoretical understanding, particularly in terms of convergence analysis and hyperparameter tuning, remains absent. In this paper, we characterize the convergence property of a widely used sampling-based MPC method, Model Predictive Path Integral Control (MPPI). We show that MPPI enjoys at least linear convergence rates when the optimization is quadratic, which covers time-varying LQR systems. We then extend to more general nonlinear systems. Our theoretical analysis directly leads to a novel sampling-based MPC algorithm, CoVariance-Optimal MPC (CoVO-MPC) that optimally schedules the sampling covariance to optimize the convergence rate. Empirically, CoVO-MPC significantly outperforms standard MPPI by 43-54% in both simulations and real-world quadrotor agile control tasks.

Reinforcement learning (RL) can align language models with non-differentiable reward signals, such as human preferences. However, a major challenge arises from the sparsity of these reward signals - typically, there is only one reward for the entire generation. This sparsity of rewards can lead to inefficient and unstable learning. In this paper, we introduce a novel framework leveraging the critique ability of LLMs to produce dense rewards throughout the learning process. Our approach incorporates a critic language model alongside the policy model. This critic is prompted with the task description, question, policy model's output, and environment's reward signal as input, and provides token or span-level dense rewards that reflect the quality of each segment of the output. We assess our approach on three text generation tasks: sentiment control, language model detoxification, and summarization. Experimental results show that incorporating artificial dense rewards in training yields consistent performance gains over the PPO baseline with holistic rewards. Furthermore, in a setting where the same model serves as both policy and critic, we demonstrate that "self-critique" rewards also boost learning efficiency.

Despite the impressive capabilities of Deep Reinforcement Learning (DRL) agents in many challenging scenarios, their black-box decision-making process significantly limits their deployment in safety-sensitive domains. Several previous self-interpretable works focus on revealing the critical states of the agent's decision. However, they cannot pinpoint the error-prone states. To address this issue, we propose a novel self-interpretable structure, named Backbone Extract Tree (BET), to better explain the agent's behavior by identify the error-prone states. At a high level, BET hypothesizes that states in which the agent consistently executes uniform decisions exhibit a reduced propensity for errors. To effectively model this phenomenon, BET expresses these states within neighborhoods, each defined by a curated set of representative states. Therefore, states positioned at a greater distance from these representative benchmarks are more prone to error. We evaluate BET in various popular RL environments and show its superiority over existing self-interpretable models in terms of explanation fidelity. Furthermore, we demonstrate a use case for providing explanations for the agents in StarCraft II, a sophisticated multi-agent cooperative game. To the best of our knowledge, we are the first to explain such a complex scenarios using a fully transparent structure.

This paper delves into applying reinforcement learning (RL) in strategy games, particularly those characterized by parity challenges, as seen in specific positions of Go and Chess and a broader range of impartial games. We propose a simulated learning process, structured within a curriculum learning framework and augmented with noisy labels, to mirror the intricacies of self-play learning scenarios. This approach thoroughly analyses how neural networks (NNs) adapt and evolve from elementary to increasingly complex game positions. Our empirical research indicates that even minimal label noise can significantly impede NNs' ability to discern effective strategies, a difficulty that intensifies with the growing complexity of the game positions. These findings underscore the urgent need for advanced methodologies in RL training, specifically tailored to counter the obstacles imposed by noisy evaluations. The development of such methodologies is crucial not only for enhancing NN proficiency in strategy games with significant parity elements but also for broadening the resilience and efficiency of RL systems across diverse and complex environments.

Imitation Learning (IL) is a widely used framework for learning imitative behavior from demonstrations. It is especially appealing for solving complex real-world tasks where handcrafting reward function is difficult, or when the goal is to mimic human expert behavior. However, the learned imitative policy can only follow the behavior in the demonstration. When applying the imitative policy, we may need to customize the policy behavior to meet different requirements coming from diverse downstream tasks. Meanwhile, we still want the customized policy to maintain its imitative nature. To this end, we formulate a new problem setting called policy customization. It defines the learning task as training a policy that inherits the characteristics of the prior policy while satisfying some additional requirements imposed by a target downstream task. We propose a novel and principled approach to interpret and determine the trade-off between the two task objectives. Specifically, we formulate the customization problem as a Markov Decision Process (MDP) with a reward function that combines 1) the inherent reward of the demonstration; and 2) the add-on reward specified by the downstream task. We propose a novel framework, Residual Q-learning, which can solve the formulated MDP by leveraging the prior policy without knowing the inherent reward or value function of the prior policy. We derive a family of residual Q-learning algorithms that can realize offline and online policy customization, and show that the proposed algorithms can effectively accomplish policy customization tasks in various environments. Demo videos and code are available on our website: https://sites.google.com/view/residualq-learning.

While AlphaZero-style reinforcement learning (RL) algorithms excel in various board games, in this paper we show that they face challenges on impartial games where players share pieces. We present a concrete example of a game - namely the children's game of Nim - and other impartial games that seem to be a stumbling block for AlphaZero-style and similar self-play reinforcement learning algorithms. Our work is built on the challenges posed by the intricacies of data distribution on the ability of neural networks to learn parity functions, exacerbated by the noisy labels issue. Our findings are consistent with recent studies showing that AlphaZero-style algorithms are vulnerable to adversarial attacks and adversarial perturbations, showing the difficulty of learning to master the games in all legal states. We show that Nim can be learned on small boards, but the learning progress of AlphaZero-style algorithms dramatically slows down when the board size increases. Intuitively, the difference between impartial games like Nim and partisan games like Chess and Go can be explained by the fact that if a small part of the board is covered for impartial games it is typically not possible to predict whether the position is won or lost as there is often zero correlation between the visible part of a partly blanked-out position and its correct evaluation. This situation starkly contrasts partisan games where a partly blanked-out board position typically provides abundant or at least non-trifle information about the value of the fully uncovered position.