Many reinforcement learning environments (e.g., Minecraft) provide only sparse rewards that indicate task completion or failure with binary values. The challenge in exploration efficiency in such environments makes it difficult for reinforcement-learning-based agents to learn complex tasks. To address this, this paper introduces an advanced learning system, named Auto MC-Reward, that leverages Large Language Models (LLMs) to automatically design dense reward functions, thereby enhancing the learning efficiency. Auto MC-Reward consists of three important components: Reward Designer, Reward Critic, and Trajectory Analyzer. Given the environment information and task descriptions, the Reward Designer first design the reward function by coding an executable Python function with predefined observation inputs. Then, our Reward Critic will be responsible for verifying the code, checking whether the code is self-consistent and free of syntax and semantic errors. Further, the Trajectory Analyzer summarizes possible failure causes and provides refinement suggestions according to collected trajectories. In the next round, Reward Designer will further refine and iterate the dense reward function based on feedback. Experiments demonstrate a significant improvement in the success rate and learning efficiency of our agents in complex tasks in Minecraft, such as obtaining diamond with the efficient ability to avoid lava, and efficiently explore trees and animals that are sparse in the plains biome.
相關內容
As a representative of a new generation of biometrics, vein identification technology offers a high level of security and convenience. Convolutional neural networks (CNNs), a prominent class of deep learning architectures, have been extensively utilized for vein identification. Since their performance and robustness are limited by small Effective Receptive Fields (e.g. 3$\times$3 kernels) and insufficient training samples, however, they are unable to extract global feature representations from vein images in an effective manner. To address these issues, we propose StarLKNet, a large kernel convolution-based palm-vein identification network, with the Mixup approach. Our StarMix learns effectively the distribution of vein features to expand samples. To enable CNNs to capture comprehensive feature representations from palm-vein images, we explored the effect of convolutional kernel size on the performance of palm-vein identification networks and designed LaKNet, a network leveraging large kernel convolution and gating mechanism. In light of the current state of knowledge, this represents an inaugural instance of the deployment of a CNN with large kernels in the domain of vein identification. Extensive experiments were conducted to validate the performance of StarLKNet on two public palm-vein datasets. The results demonstrated that StarMix provided superior augmentation, and LakNet exhibited more stable performance gains compared to mainstream approaches, resulting in the highest recognition accuracy and lowest identification error.
Thanks to the explosive developments of data-driven learning methodologies recently, reinforcement learning (RL) emerges as a promising solution to address the legged locomotion problem in robotics. In this manuscript, we propose a novel concurrent teacher-student reinforcement learning architecture for legged locomotion over challenging terrains, based only on proprioceptive measurements in real-world deployment. Different from convectional teacher-student architecture that trains the teacher policy via RL and transfers the knowledge to the student policy through supervised learning, our proposed architecture trains teacher and student policy networks concurrently under the reinforcement learning paradigm. To achieve this, we develop a new training scheme based on conventional proximal policy gradient (PPO) method to accommodate the interaction between teacher policy network and student policy network. The effectiveness of the proposed architecture as well as the new training scheme is demonstrated through extensive indoor and outdoor experiments on quadrupedal robots and point-foot bipedal robot, showcasing robust locomotion over challenging terrains and improved performance compared to two-stage training methods.
Before deploying outputs from foundation models in high-stakes tasks, it is imperative to ensure that they align with human values. For instance, in radiology report generation, reports generated by a vision-language model must align with human evaluations before their use in medical decision-making. This paper presents Conformal Alignment, a general framework for identifying units whose outputs meet a user-specified alignment criterion. It is guaranteed that on average, a prescribed fraction of selected units indeed meet the alignment criterion, regardless of the foundation model or the data distribution. Given any pre-trained model and new units with model-generated outputs, Conformal Alignment leverages a set of reference data with ground-truth alignment status to train an alignment predictor. It then selects new units whose predicted alignment scores surpass a data-dependent threshold, certifying their corresponding outputs as trustworthy. Through applications to question answering and radiology report generation, we demonstrate that our method is able to accurately identify units with trustworthy outputs via lightweight training over a moderate amount of reference data. En route, we investigate the informativeness of various features in alignment prediction and combine them with standard models to construct the alignment predictor.
Value function factorization methods are commonly used in cooperative multi-agent reinforcement learning, with QMIX receiving significant attention. Many QMIX-based methods introduce monotonicity constraints between the joint action value and individual action values to achieve decentralized execution. However, such constraints limit the representation capacity of value factorization, restricting the joint action values it can represent and hindering the learning of the optimal policy. To address this challenge, we propose the Potentially Optimal joint actions Weighted QMIX (POWQMIX) algorithm, which recognizes the potentially optimal joint actions and assigns higher weights to the corresponding losses of these joint actions during training. We theoretically prove that with such a weighted training approach the optimal policy is guaranteed to be recovered. Experiments in matrix games, predator-prey, and StarCraft II Multi-Agent Challenge environments demonstrate that our algorithm outperforms the state-of-the-art value-based multi-agent reinforcement learning methods.
Adversarial inverse reinforcement learning (AIRL) stands as a cornerstone approach in imitation learning, yet it faces criticisms from prior studies. In this paper, we rethink AIRL and respond to these criticisms. Criticism 1 lies in Inadequate Policy Imitation. We show that substituting the built-in algorithm with soft actor-critic (SAC) during policy updating (requires multi-iterations) significantly enhances the efficiency of policy imitation. Criticism 2 lies in Limited Performance in Transferable Reward Recovery Despite SAC Integration. While we find that SAC indeed exhibits a significant improvement in policy imitation, it introduces drawbacks to transferable reward recovery. We prove that the SAC algorithm itself is not feasible to disentangle the reward function comprehensively during the AIRL training process, and propose a hybrid framework, PPO-AIRL + SAC, for a satisfactory transfer effect. Criticism 3 lies in Unsatisfactory Proof from the Perspective of Potential Equilibrium. We reanalyze it from an algebraic theory perspective.
Many aging individuals encounter challenges in effectively tracking their dietary intake, exacerbating their susceptibility to nutrition-related health complications. Self-reporting methods are often inaccurate and suffer from substantial bias; however, leveraging intelligent prediction methods can automate and enhance precision in this process. Recent work has explored using computer vision prediction systems to predict nutritional information from food images. Still, these methods are often tailored to specific situations, require other inputs in addition to a food image, or do not provide comprehensive nutritional information. This paper aims to enhance the efficacy of dietary intake estimation by leveraging various neural network architectures to directly predict a meal's nutritional content from its image. Through comprehensive experimentation and evaluation, we present NutritionVerse-Direct, a model utilizing a vision transformer base architecture with three fully connected layers that lead to five regression heads predicting calories (kcal), mass (g), protein (g), fat (g), and carbohydrates (g) present in a meal. NutritionVerse-Direct yields a combined mean average error score on the NutritionVerse-Real dataset of 412.6, an improvement of 25.5% over the Inception-ResNet model, demonstrating its potential for improving dietary intake estimation accuracy.
Developing accurate machine learning models for oncology requires large-scale, high-quality multimodal datasets. However, creating such datasets remains challenging due to the complexity and heterogeneity of medical data. To address this challenge, we introduce HoneyBee, a scalable modular framework for building multimodal oncology datasets that leverages foundational models to generate representative embeddings. HoneyBee integrates various data modalities, including clinical records, imaging data, and patient outcomes. It employs data preprocessing techniques and transformer-based architectures to generate embeddings that capture the essential features and relationships within the raw medical data. The generated embeddings are stored in a structured format using Hugging Face datasets and PyTorch dataloaders for accessibility. Vector databases enable efficient querying and retrieval for machine learning applications. We demonstrate the effectiveness of HoneyBee through experiments assessing the quality and representativeness of the embeddings. The framework is designed to be extensible to other medical domains and aims to accelerate oncology research by providing high-quality, machine learning-ready datasets. HoneyBee is an ongoing open-source effort, and the code, datasets, and models are available at the project repository.
This work investigates the modeling and optimization of beyond diagonal reconfigurable intelligent surface (BD-RIS), which generalizes conventional RIS with diagonal phase shift matrices and provides additional flexibility for manipulating wireless channels, in wideband communication systems. Specifically, we start from the signal modeling of the BD-RIS-aided orthogonal frequency division multiplexing (OFDM) system, which bridges the time-domain and frequency-domain channels, and explicitly shows the frequency dependence of the BD-RIS response. We next characterize the frequency dependence of the BD-RIS response based on circuit models. Benefiting from the admittance parameter analysis, we model individually each tunable admittance component of BD-RIS and derive an approximated linear expression with respect to the frequency of the transmit signals. With the proposed signal model for the BD-RIS-aided OFDM system and the frequency-dependent BD-RIS model, we propose algorithms to optimize the BD-RIS and the power allocation at the transmitter to maximize the average rate for a BD-RIS-aided OFDM system. Finally, simulation results show that BD-RIS outperforms conventional RIS in the OFDM system. More importantly, the impact of wideband modeling of BD-RIS on the system performance becomes more significant as the circuit complexity of BD-RIS architectures increases.
To solve complex tasks under resource constraints, reinforcement learning (RL) agents need to be simple, efficient, and scalable, addressing (1) large state spaces and (2) the continuous accumulation of interaction data. We propose HyperAgent, an RL framework featuring the hypermodel and index sampling schemes that enable computation-efficient incremental approximation for the posteriors associated with general value functions without the need for conjugacy, and data-efficient action selection. Implementing HyperAgent is straightforward, requiring only one additional module beyond what is necessary for Double-DQN. HyperAgent stands out as the first method to offer robust performance in large-scale deep RL benchmarks while achieving provably scalable per-step computational complexity and attaining sublinear regret under tabular assumptions. HyperAgent can solve Deep Sea hard exploration problems with episodes that optimally scale with problem size and exhibits significant efficiency gains in both data and computation under the Atari benchmark. The core of our theoretical analysis is the sequential posterior approximation argument, enabled by the first analytical tool for sequential random projection -- a non-trivial martingale extension of the Johnson-Lindenstrauss. This work bridges the theoretical and practical realms of RL, establishing a new benchmark for RL algorithm design.
Feedback-driven optimization, such as traditional machine learning training, is a static process that lacks real-time adaptability of hyperparameters. Tuning solutions for optimization require trial and error paired with checkpointing and schedulers, in many cases feedback from the algorithm is overlooked. Adjusting hyperparameters during optimization usually requires the program to be restarted, wasting utilization and time, while placing unnecessary strain on memory and processors. We present LiveTune, a novel framework allowing real-time parameter adjustment of optimization loops through LiveVariables. Live Variables allow for continuous feedback-driven optimization by storing parameters on designated ports on the system, allowing them to be dynamically adjusted. Extensive evaluations of our framework on standard machine learning training pipelines show saving up to 60 seconds and 5.4 Kilojoules of energy per hyperparameter change. We also show the feasibility and value of LiveTune in a reinforcement learning application where the users change the dynamics of the reward structure while the agent is learning showing 5x improvement over the baseline. Finally, we outline a fully automated workflow to provide end-to-end, unsupervised feedback-driven optimization.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191