PyTorchKR Papers

최신 논문

총 3,220편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2404.04682공식 코드 · Papers with Code

Compositional Conservatism: A Transductive Approach in Offline Reinforcement Learning

Offline reinforcement learning (RL) is a compelling framework for learning optimal policies from past experiences without additional interaction with the environment. Nevertheless, offline RL inevitably faces the problem of distributional shifts, where the states and actions encountered during policy execution may not be in the training dataset distribution. A common solution involves incorporating conservatism into the policy or the value function to safeguard against uncertainties and unknowns. In this work, we focus on achieving the same objectives of conservatism but from a different perspective. We propose COmpositional COnservatism with Anchor-seeking (COCOA) for offline RL, an approach that pursues conservatism in a compositional manner on top of the transductive reparameterization (Netanyahu et al., 2023), which decomposes the input variable (the state in our case) into an anchor and its difference from the original input. Our COCOA seeks both in-distribution anchors and differences by utilizing the learned reverse dynamics model, encouraging conservatism in the compositional input space for the policy or value function. Such compositional conservatism is independent of and agnostic to the prevalent behavioral conservatism in offline RL. We apply COCOA to four state-of-the-art offline RL algorithms and evaluate them on the D4RL benchmark, where COCOA generally improves the performance of each algorithm. The code is available at https://github.com/runamu/compositional-conservatism.

arXiv:2404.03673공식 코드 · Papers with Code

RL for Consistency Models: Faster Reward Guided Text-to-Image Generation

Reinforcement learning (RL) has improved guided image generation with diffusion models by directly optimizing rewards that capture image quality, aesthetics, and instruction following capabilities. However, the resulting generative policies inherit the same iterative sampling process of diffusion models that causes slow generation. To overcome this limitation, consistency models proposed learning a new class of generative models that directly map noise to data, resulting in a model that can generate an image in as few as one sampling iteration. In this work, to optimize text-to-image generative models for task specific rewards and enable fast training and inference, we propose a framework for fine-tuning consistency models via RL. Our framework, called Reinforcement Learning for Consistency Model (RLCM), frames the iterative inference process of a consistency model as an RL procedure. Comparing to RL finetuned diffusion models, RLCM trains significantly faster, improves the quality of the generation measured under the reward objectives, and speeds up the inference procedure by generating high quality images with as few as two inference steps. Experimentally, we show that RLCM can adapt text-to-image consistency models to objectives that are challenging to express with prompting, such as image compressibility, and those derived from human feedback, such as aesthetic quality. Our code is available at https://rlcm.owenoertell.com.

arXiv:2404.03037공식 코드 · Papers with Code

Model-based Reinforcement Learning for Parameterized Action Spaces

We propose a novel model-based reinforcement learning algorithm -- Dynamics Learning and predictive control with Parameterized Actions (DLPA) -- for Parameterized Action Markov Decision Processes (PAMDPs). The agent learns a parameterized-action-conditioned dynamics model and plans with a modified Model Predictive Path Integral control. We theoretically quantify the difference between the generated trajectory and the optimal trajectory during planning in terms of the value they achieved through the lens of Lipschitz Continuity. Our empirical results on several standard benchmarks show that our algorithm achieves superior sample efficiency and asymptotic performance than state-of-the-art PAMDP methods.

arXiv:2404.02545공식 코드 · Papers with Code

Grid-Mapping Pseudo-Count Constraint for Offline Reinforcement Learning

Offline reinforcement learning learns from a static dataset without interacting with environments, which ensures security and thus owns a good application prospect. However, directly applying naive reinforcement learning algorithm usually fails in an offline environment due to inaccurate Q value approximation caused by out-of-distribution (OOD) state-actions. It is an effective way to solve this problem by penalizing the Q-value of OOD state-actions. Among the methods of punishing OOD state-actions, count-based methods have achieved good results in discrete domains in a simple form. Inspired by it, a novel pseudo-count method for continuous domains called Grid-Mapping Pseudo-Count method (GPC) is proposed by extending the count-based method from discrete to continuous domains. Firstly, the continuous state and action space are mapped to discrete space using Grid-Mapping, then the Q-values of OOD state-actions are constrained through pseudo-count. Secondly, the theoretical proof is given to show that GPC can obtain appropriate uncertainty constraints under fewer assumptions than other pseudo-count methods. Thirdly, GPC is combined with Soft Actor-Critic algorithm (SAC) to get a new algorithm called GPC-SAC. Lastly, experiments on D4RL datasets are given to show that GPC-SAC has better performance and less computational cost than other algorithms that constrain the Q-value.

arXiv:2404.00978공식 코드 · Papers with Code

Prior Constraints-based Reward Model Training for Aligning Large Language Models

Reinforcement learning with human feedback for aligning large language models (LLMs) trains a reward model typically using ranking loss with comparison pairs.However, the training procedure suffers from an inherent problem: the uncontrolled scaling of reward scores during reinforcement learning due to the lack of constraints while training the reward model.This paper proposes a Prior Constraints-based Reward Model (namely PCRM) training method to mitigate this problem. PCRM incorporates prior constraints, specifically, length ratio and cosine similarity between outputs of each comparison pair, during reward model training to regulate optimization magnitude and control score margins. We comprehensively evaluate PCRM by examining its rank correlation with human preferences and its effectiveness in aligning LLMs via RL. Experimental results demonstrate that PCRM significantly improves alignment performance by effectively constraining reward score scaling. As another bonus, our method is easily integrated into arbitrary rank-based alignment methods, such as direct preference optimization, and can yield consistent improvement.

arXiv:2404.01617공식 코드 · Papers with Code

Designing Network Algorithms via Large Language Models

We introduce NADA, the first framework to autonomously design network algorithms by leveraging the generative capabilities of large language models (LLMs). Starting with an existing algorithm implementation, NADA enables LLMs to create a wide variety of alternative designs in the form of code blocks. It then efficiently identifies the top-performing designs through a series of filtering techniques, minimizing the need for full-scale evaluations and significantly reducing computational costs. Using adaptive bitrate (ABR) streaming as a case study, we demonstrate that NADA produces novel ABR algorithms -- previously unknown to human developers -- that consistently outperform the original algorithm in diverse network environments, including broadband, satellite, 4G, and 5G.

arXiv:2404.01220공식 코드 · Papers with Code

Entity-Centric Reinforcement Learning for Object Manipulation from Pixels

Manipulating objects is a hallmark of human intelligence, and an important task in domains such as robotics. In principle, Reinforcement Learning (RL) offers a general approach to learn object manipulation. In practice, however, domains with more than a few objects are difficult for RL agents due to the curse of dimensionality, especially when learning from raw image observations. In this work we propose a structured approach for visual RL that is suitable for representing multiple objects and their interaction, and use it to learn goal-conditioned manipulation of several objects. Key to our method is the ability to handle goals with dependencies between the objects (e.g., moving objects in a certain order). We further relate our architecture to the generalization capability of the trained agent, based on a theoretical result for compositional generalization, and demonstrate agents that learn with 3 objects but generalize to similar tasks with over 10 objects. Videos and code are available on the project website: https://sites.google.com/view/entity-centric-rl

arXiv:2403.19925공식 코드 · Papers with Code

Decision Mamba: Reinforcement Learning via Sequence Modeling with Selective State Spaces

Decision Transformer, a promising approach that applies Transformer architectures to reinforcement learning, relies on causal self-attention to model sequences of states, actions, and rewards. While this method has shown competitive results, this paper investigates the integration of the Mamba framework, known for its advanced capabilities in efficient and effective sequence modeling, into the Decision Transformer architecture, focusing on the potential performance enhancements in sequential decision-making tasks. Our study systematically evaluates this integration by conducting a series of experiments across various decision-making environments, comparing the modified Decision Transformer, Decision Mamba, with its traditional counterpart. This work contributes to the advancement of sequential decision-making models, suggesting that the architecture and training methodology of neural networks can significantly impact their performance in complex tasks, and highlighting the potential of Mamba as a valuable tool for improving the efficacy of Transformer-based models in reinforcement learning scenarios.

arXiv:2403.19839공식 코드 · Papers with Code

The New Agronomists: Language Models are Experts in Crop Management

Crop management plays a crucial role in determining crop yield, economic profitability, and environmental sustainability. Despite the availability of management guidelines, optimizing these practices remains a complex and multifaceted challenge. In response, previous studies have explored using reinforcement learning with crop simulators, typically employing simple neural-network-based reinforcement learning (RL) agents. Building on this foundation, this paper introduces a more advanced intelligent crop management system. This system uniquely combines RL, a language model (LM), and crop simulations facilitated by the Decision Support System for Agrotechnology Transfer (DSSAT). We utilize deep RL, specifically a deep Q-network, to train management policies that process numerous state variables from the simulator as observations. A novel aspect of our approach is the conversion of these state variables into more informative language, facilitating the language model's capacity to understand states and explore optimal management practices. The empirical results reveal that the LM exhibits superior learning capabilities. Through simulation experiments with maize crops in Florida (US) and Zaragoza (Spain), the LM not only achieves state-of-the-art performance under various evaluation metrics but also demonstrates a remarkable improvement of over 49\% in economic profit, coupled with reduced environmental impact when compared to baseline methods. Our code is available at \url{https://github.com/jingwu6/LM_AG}.

arXiv:2403.20109공식 코드 · Papers with Code

Mol-AIR: Molecular Reinforcement Learning with Adaptive Intrinsic Rewards for Goal-directed Molecular Generation

Optimizing techniques for discovering molecular structures with desired properties is crucial in artificial intelligence(AI)-based drug discovery. Combining deep generative models with reinforcement learning has emerged as an effective strategy for generating molecules with specific properties. Despite its potential, this approach is ineffective in exploring the vast chemical space and optimizing particular chemical properties. To overcome these limitations, we present Mol-AIR, a reinforcement learning-based framework using adaptive intrinsic rewards for effective goal-directed molecular generation. Mol-AIR leverages the strengths of both history-based and learning-based intrinsic rewards by exploiting random distillation network and counting-based strategies. In benchmark tests, Mol-AIR demonstrates superior performance over existing approaches in generating molecules with desired properties without any prior knowledge, including penalized LogP, QED, and celecoxib similarity. We believe that Mol-AIR represents a significant advancement in drug discovery, offering a more efficient path to discovering novel therapeutics.

arXiv:2403.19024공식 코드 · Papers with Code

Exploiting Symmetry in Dynamics for Model-Based Reinforcement Learning with Asymmetric Rewards

Recent work in reinforcement learning has leveraged symmetries in the model to improve sample efficiency in training a policy. A commonly used simplifying assumption is that the dynamics and reward both exhibit the same symmetry; however, in many real-world environments, the dynamical model exhibits symmetry independent of the reward model. In this paper, we assume only the dynamics exhibit symmetry, extending the scope of problems in reinforcement learning and learning in control theory to which symmetry techniques can be applied. We use Cartan's moving frame method to introduce a technique for learning dynamics that, by construction, exhibit specified symmetries. Numerical experiments demonstrate that the proposed method learns a more accurate dynamical model

arXiv:2403.18219공식 코드 · Papers with Code

From Two-Dimensional to Three-Dimensional Environment with Q-Learning: Modeling Autonomous Navigation with Reinforcement Learning and no Libraries

Reinforcement learning (RL) algorithms have become indispensable tools in artificial intelligence, empowering agents to acquire optimal decision-making policies through interactions with their environment and feedback mechanisms. This study explores the performance of RL agents in both two-dimensional (2D) and three-dimensional (3D) environments, aiming to research the dynamics of learning across different spatial dimensions. A key aspect of this investigation is the absence of pre-made libraries for learning, with the algorithm developed exclusively through computational mathematics. The methodological framework centers on RL principles, employing a Q-learning agent class and distinct environment classes tailored to each spatial dimension. The research aims to address the question: How do reinforcement learning agents adapt and perform in environments of varying spatial dimensions, particularly in 2D and 3D settings? Through empirical analysis, the study evaluates agents' learning trajectories and adaptation processes, revealing insights into the efficacy of RL algorithms in navigating complex, multi-dimensional spaces. Reflections on the findings prompt considerations for future research, particularly in understanding the dynamics of learning in higher-dimensional environments.

arXiv:2403.17456공식 코드 · Papers with Code

Imitating Cost-Constrained Behaviors in Reinforcement Learning

Complex planning and scheduling problems have long been solved using various optimization or heuristic approaches. In recent years, imitation learning that aims to learn from expert demonstrations has been proposed as a viable alternative to solving these problems. Generally speaking, imitation learning is designed to learn either the reward (or preference) model or directly the behavioral policy by observing the behavior of an expert. Existing work in imitation learning and inverse reinforcement learning has focused on imitation primarily in unconstrained settings (e.g., no limit on fuel consumed by the vehicle). However, in many real-world domains, the behavior of an expert is governed not only by reward (or preference) but also by constraints. For instance, decisions on self-driving delivery vehicles are dependent not only on the route preferences/rewards (depending on past demand data) but also on the fuel in the vehicle and the time available. In such problems, imitation learning is challenging as decisions are not only dictated by the reward model but are also dependent on a cost-constrained model. In this paper, we provide multiple methods that match expert distributions in the presence of trajectory cost constraints through (a) Lagrangian-based method; (b) Meta-gradients to find a good trade-off between expected return and minimizing constraint violation; and (c) Cost-violation-based alternating gradient. We empirically show that leading imitation learning approaches imitate cost-constrained behaviors poorly and our meta-gradient-based approach achieves the best performance.

arXiv:2403.17637공식 코드 · Papers with Code

PeersimGym: An Environment for Solving the Task Offloading Problem with Reinforcement Learning

Task offloading, crucial for balancing computational loads across devices in networks such as the Internet of Things, poses significant optimization challenges, including minimizing latency and energy usage under strict communication and storage constraints. While traditional optimization falls short in scalability; and heuristic approaches lack in achieving optimal outcomes, Reinforcement Learning (RL) offers a promising avenue by enabling the learning of optimal offloading strategies through iterative interactions. However, the efficacy of RL hinges on access to rich datasets and custom-tailored, realistic training environments. To address this, we introduce PeersimGym, an open-source, customizable simulation environment tailored for developing and optimizing task offloading strategies within computational networks. PeersimGym supports a wide range of network topologies and computational constraints and integrates a \textit{PettingZoo}-based interface for RL agent deployment in both solo and multi-agent setups. Furthermore, we demonstrate the utility of the environment through experiments with Deep Reinforcement Learning agents, showcasing the potential of RL-based approaches to significantly enhance offloading strategies in distributed computing settings. PeersimGym thus bridges the gap between theoretical RL models and their practical applications, paving the way for advancements in efficient task offloading methodologies.

arXiv:2403.16391공식 코드 · Papers with Code

Physics-informed RL for Maximal Safety Probability Estimation

Accurate risk quantification and reachability analysis are crucial for safe control and learning, but sampling from rare events, risky states, or long-term trajectories can be prohibitively costly. Motivated by this, we study how to estimate the long-term safety probability of maximally safe actions without sufficient coverage of samples from risky states and long-term trajectories. The use of maximal safety probability in control and learning is expected to avoid conservative behaviors due to over-approximation of risk. Here, we first show that long-term safety probability, which is multiplicative in time, can be converted into additive costs and be solved using standard reinforcement learning methods. We then derive this probability as solutions of partial differential equations (PDEs) and propose Physics-Informed Reinforcement Learning (PIRL) algorithm. The proposed method can learn using sparse rewards because the physics constraints help propagate risk information through neighbors. This suggests that, for the purpose of extracting more information for efficient learning, physics constraints can serve as an alternative to reward shaping. The proposed method can also estimate long-term risk using short-term samples and deduce the risk of unsampled states. This feature is in stark contrast with the unconstrained deep RL that demands sufficient data coverage. These merits of the proposed method are demonstrated in numerical simulation.

arXiv:2403.16649공식 코드 · Papers with Code

CLHA: A Simple yet Effective Contrastive Learning Framework for Human Alignment

Reinforcement learning from human feedback (RLHF) is a crucial technique in aligning large language models (LLMs) with human preferences, ensuring these LLMs behave in beneficial and comprehensible ways to users. However, a longstanding challenge in human alignment techniques based on reinforcement learning lies in their inherent complexity and difficulty in training. To address this challenge, we present a simple yet effective Contrastive Learning Framework for Human Alignment (CLHA) to align LLMs with human preferences directly. CLHA employs a novel rescoring strategy to evaluate the noise within the data by considering its inherent quality and dynamically adjusting the training process. Simultaneously, CLHA utilizes pairwise contrastive loss and adaptive supervised fine-tuning loss to adaptively modify the likelihood of generating responses, ensuring enhanced alignment with human preferences. Using advanced methods, CLHA surpasses other algorithms, showcasing superior performance in terms of reward model scores, automatic evaluations, and human assessments on the widely used ``Helpful and Harmless'' dataset.

arXiv:2403.14405공식 코드 · Papers with Code

A reinforcement learning guided hybrid evolutionary algorithm for the latency location routing problem

The latency location routing problem integrates the facility location problem and the multi-depot cumulative capacitated vehicle routing problem. This problem involves making simultaneous decisions about depot locations and vehicle routes to serve customers while aiming to minimize the sum of waiting (arriving) times for all customers. To address this computationally challenging problem, we propose a reinforcement learning guided hybrid evolutionary algorithm following the framework of the memetic algorithm. The proposed algorithm relies on a diversity-enhanced multi-parent edge assembly crossover to build promising offspring and a reinforcement learning guided variable neighborhood descent to determine the exploration order of multiple neighborhoods. Additionally, strategic oscillation is used to achieve a balanced exploration of both feasible and infeasible solutions. The competitiveness of the algorithm against state-of-the-art methods is demonstrated by experimental results on the three sets of 76 popular instances, including 51 improved best solutions (new upper bounds) for the 59 instances with unknown optima and equal best results for the remaining instances. We also conduct additional experiments to shed light on the key components of the algorithm.

arXiv:2403.14273공식 코드 · Papers with Code

Reactor Optimization Benchmark by Reinforcement Learning

Neutronic calculations for reactors are a daunting task when using Monte Carlo (MC) methods. As high-performance computing has advanced, the simulation of a reactor is nowadays more readily done, but design and optimization with multiple parameters is still a computational challenge. MC transport simulations, coupled with machine learning techniques, offer promising avenues for enhancing the efficiency and effectiveness of nuclear reactor optimization. This paper introduces a novel benchmark problem within the OpenNeoMC framework designed specifically for reinforcement learning. The benchmark involves optimizing a unit cell of a research reactor with two varying parameters (fuel density and water spacing) to maximize neutron flux while maintaining reactor criticality. The test case features distinct local optima, representing different physical regimes, thus posing a challenge for learning algorithms. Through extensive simulations utilizing evolutionary and neuroevolutionary algorithms, we demonstrate the effectiveness of reinforcement learning in navigating complex optimization landscapes with strict constraints. Furthermore, we propose acceleration techniques within the OpenNeoMC framework, including model updating and cross-section usage by RAM utilization, to expedite simulation times. Our findings emphasize the importance of machine learning integration in reactor optimization and contribute to advancing methodologies for addressing intricate optimization challenges in nuclear engineering. The sources of this work are available at our GitHub repository: https://github.com/Scientific-Computing-Lab-NRCN/RLOpenNeoMC

arXiv:2403.13765공식 코드 · Papers with Code

Towards Principled Representation Learning from Videos for Reinforcement Learning

We study pre-training representations for decision-making using video data, which is abundantly available for tasks such as game agents and software testing. Even though significant empirical advances have been made on this problem, a theoretical understanding remains absent. We initiate the theoretical investigation into principled approaches for representation learning and focus on learning the latent state representations of the underlying MDP using video data. We study two types of settings: one where there is iid noise in the observation, and a more challenging setting where there is also the presence of exogenous noise, which is non-iid noise that is temporally correlated, such as the motion of people or cars in the background. We study three commonly used approaches: autoencoding, temporal contrastive learning, and forward modeling. We prove upper bounds for temporal contrastive learning and forward modeling in the presence of only iid noise. We show that these approaches can learn the latent state and use it to do efficient downstream RL with polynomial sample complexity. When exogenous noise is also present, we establish a lower bound result showing that the sample complexity of learning from video data can be exponentially worse than learning from action-labeled trajectory data. This partially explains why reinforcement learning with video pre-training is hard. We evaluate these representational learning methods in two visual domains, yielding results that are consistent with our theoretical findings.

arXiv:2403.12847공식 코드 · Papers with Code

Policy Bifurcation in Safe Reinforcement Learning

Safe reinforcement learning (RL) offers advanced solutions to constrained optimal control problems. Existing studies in safe RL implicitly assume continuity in policy functions, where policies map states to actions in a smooth, uninterrupted manner; however, our research finds that in some scenarios, the feasible policy should be discontinuous or multi-valued, interpolating between discontinuous local optima can inevitably lead to constraint violations. We are the first to identify the generating mechanism of such a phenomenon, and employ topological analysis to rigorously prove the existence of policy bifurcation in safe RL, which corresponds to the contractibility of the reachable tuple. Our theorem reveals that in scenarios where the obstacle-free state space is non-simply connected, a feasible policy is required to be bifurcated, meaning its output action needs to change abruptly in response to the varying state. To train such a bifurcated policy, we propose a safe RL algorithm called multimodal policy optimization (MUPO), which utilizes a Gaussian mixture distribution as the policy output. The bifurcated behavior can be achieved by selecting the Gaussian component with the highest mixing coefficient. Besides, MUPO also integrates spectral normalization and forward KL divergence to enhance the policy's capability of exploring different modes. Experiments with vehicle control tasks show that our algorithm successfully learns the bifurcated policy and ensures satisfying safety, while a continuous policy suffers from inevitable constraint violations.