Artificial Intelligence (AI) systems have made remarkable progress, attaining super-human performance across various domains. This presents us with an opportunity to further human knowledge and improve human expert performance by leveraging the hidden knowledge encoded within these highly performant AI systems. Yet, this knowledge is often hard to extract, and may be hard to understand or learn from. Here, we show that this is possible by proposing a new method that allows us to extract new chess concepts in AlphaZero, an AI system that mastered the game of chess via self-play without human supervision. Our analysis indicates that AlphaZero may encode knowledge that extends beyond the existing human knowledge, but knowledge that is ultimately not beyond human grasp, and can be successfully learned from. In a human study, we show that these concepts are learnable by top human experts, as four top chess grandmasters show improvements in solving the presented concept prototype positions. This marks an important first milestone in advancing the frontier of human knowledge by leveraging AI; a development that could bear profound implications and help us shape how we interact with AI systems across many AI applications.

Reinforcement Learning (RL) is notoriously data-inefficient, which makes training on a real robot difficult. While model-based RL algorithms (world models) improve data-efficiency to some extent, they still require hours or days of interaction to learn skills. Recently, offline RL has been proposed as a framework for training RL policies on pre-existing datasets without any online interaction. However, constraining an algorithm to a fixed dataset induces a state-action distribution shift between training and inference, and limits its applicability to new tasks. In this work, we seek to get the best of both worlds: we consider the problem of pretraining a world model with offline data collected on a real robot, and then finetuning the model on online data collected by planning with the learned model. To mitigate extrapolation errors during online interaction, we propose to regularize the planner at test-time by balancing estimated returns and (epistemic) model uncertainty. We evaluate our method on a variety of visuo-motor control tasks in simulation and on a real robot, and find that our method enables few-shot finetuning to seen and unseen tasks even when offline data is limited. Videos, code, and data are available at https://yunhaifeng.com/FOWM .

Different methods are used for a mobile robot to go to a specific target location. These methods work in different ways for online and offline scenarios. In the offline scenario, an environment map is created once, and fixed path planning is made on this map to reach the target. Path planning algorithms such as A* and RRT (Rapidly-Exploring Random Tree) are the examples of offline methods. The most obvious situation here is the need to re-plan the path for changing conditions of the loaded map. On the other hand, in the online scenario, the robot moves dynamically to a given target without using a map by using the perceived data coming from the sensors. Approaches such as SFM (Social Force Model) are used in online systems. However, these methods suffer from the requirement of a lot of dynamic sensing data. Thus, it can be said that the need for re-planning and mapping in offline systems and various system design requirements in online systems are the subjects that focus on autonomous mobile robot research. Recently, deep neural network powered Q-Learning methods are used as an emerging solution to the aforementioned problems in mobile robot navigation. In this study, machine learning algorithms with deep Q-Learning (DQN) and Deep DQN architectures, are evaluated for the solution of the problems presented above to realize path planning of an autonomous mobile robot to avoid obstacles.

The self-attention mechanism in the transformer architecture is capable of capturing long-range dependencies and it is the main reason behind its effectiveness in processing sequential data. Nevertheless, despite their success, transformers have two significant drawbacks that still limit their broader applicability: (1) In order to remember past information, the self-attention mechanism requires access to the whole history to be provided as context. (2) The inference cost in transformers is expensive. In this paper we introduce recurrent alternatives to the transformer self-attention mechanism that offer a context-independent inference cost, leverage long-range dependencies effectively, and perform well in practice. We evaluate our approaches in reinforcement learning problems where the aforementioned computational limitations make the application of transformers nearly infeasible. We quantify the impact of the different components of our architecture in a diagnostic environment and assess performance gains in 2D and 3D pixel-based partially-observable environments. When compared to a state-of-the-art architecture, GTrXL, inference in our approach is at least 40% cheaper while reducing memory use in more than 50%. Our approach either performs similarly or better than GTrXL, improving more than 37% upon GTrXL performance on harder tasks.

High-frequency trading is prevalent, where automated decisions must be made quickly to take advantage of price imbalances and patterns in price action that forecast near-future movements. While many algorithms have been explored and tested, analytical methods fail to harness the whole nature of the market environment by focusing on a limited domain. With the evergrowing machine learning field, many large-scale end-to-end studies on raw data have been successfully employed to increase the domain scope for profitable trading but are very difficult to replicate. Combining deep learning on the order books with reinforcement learning is one way of breaking down large-scale end-to-end learning into more manageable and lightweight components for reproducibility, suitable for retail trading. The following work focuses on forecasting returns across multiple horizons using order flow imbalance and training three temporal-difference learning models for five financial instruments to provide trading signals. The instruments used are two foreign exchange pairs (GBPUSD and EURUSD), two indices (DE40 and FTSE100), and one commodity (XAUUSD). The performances of these 15 agents are evaluated through backtesting simulation, and successful models proceed through to forward testing on a retail trading platform. The results prove potential but require further minimal modifications for consistently profitable trading to fully handle retail trading costs, slippage, and spread fluctuation.

Owing to the recent developments in Generative Artificial Intelligence (GenAI) and Large Language Models (LLM), conversational agents are becoming increasingly popular and accepted. They provide a human touch by interacting in ways familiar to us and by providing support as virtual companions. Therefore, it is important to understand the user's emotions in order to respond considerately. Compared to the standard problem of emotion recognition, conversational agents face an additional constraint in that recognition must be real-time. Studies on model architectures using audio, visual, and textual modalities have mainly focused on emotion classification using full video sequences that do not provide online features. In this work, we present a novel paradigm for contextualized Emotion Recognition using Graph Convolutional Network with Reinforcement Learning (conER-GRL). Conversations are partitioned into smaller groups of utterances for effective extraction of contextual information. The system uses Gated Recurrent Units (GRU) to extract multimodal features from these groups of utterances. More importantly, Graph Convolutional Networks (GCN) and Reinforcement Learning (RL) agents are cascade trained to capture the complex dependencies of emotion features in interactive scenarios. Comparing the results of the conER-GRL model with other state-of-the-art models on the benchmark dataset IEMOCAP demonstrates the advantageous capabilities of the conER-GRL architecture in recognizing emotions in real-time from multimodal conversational signals.

The Mean-Field approximation is a tractable approach for studying large population dynamics. However, its assumption on homogeneity and universal connections among all agents limits its applicability in many real-world scenarios. Multi-Population Mean-Field Game (MP-MFG) models have been introduced in the literature to address these limitations. When the underlying Stochastic Block Model is known, we show that a Policy Mirror Ascent algorithm finds the MP-MFG Nash Equilibrium. In more realistic scenarios where the block model is unknown, we propose a re-sampling scheme from a graphon integrated with the finite N-player MP-MFG model. We develop a novel learning framework based on a Graphon Game with Re-Sampling (GGR-S) model, which captures the complex network structures of agents' connections. We analyze GGR-S dynamics and establish the convergence to dynamics of MP-MFG. Leveraging this result, we propose an efficient sample-based N-player Reinforcement Learning algorithm for GGR-S without population manipulation, and provide a rigorous convergence analysis with finite sample guarantee.

Deep Reinforcement Learning (DRL) is widely used to optimize the performance of multi-UAV networks. However, the training of DRL relies on the frequent interactions between the UAVs and the environment, which consumes lots of energy due to the flying and communication of UAVs in practical experiments. Inspired by the growing digital twin (DT) technology, which can simulate the performance of algorithms in the digital space constructed by coping features of the physical space, the DT is introduced to reduce the costs of practical training, e.g., energy and hardware purchases. Different from previous DT-assisted works with an assumption of perfect reflecting real physics by virtual digital, we consider an imperfect DT model with deviations for assisting the training of multi-UAV networks. Remarkably, to trade off the training cost, DT construction cost, and the impact of deviations of DT on training, the natural and virtually generated UAV mixing deployment method is proposed. Two cascade neural networks (NN) are used to optimize the joint number of virtually generated UAVs, the DT construction cost, and the performance of multi-UAV networks. These two NNs are trained by unsupervised and reinforcement learning, both low-cost label-free training methods. Simulation results show the training cost can significantly decrease while guaranteeing the training performance. This implies that an efficient decision can be made with imperfect DTs in multi-UAV networks.

This paper provides a theoretical understanding of Deep Q-Network (DQN) with the $\varepsilon$-greedy exploration in deep reinforcement learning. Despite the tremendous empirical achievement of the DQN, its theoretical characterization remains underexplored. First, the exploration strategy is either impractical or ignored in the existing analysis. Second, in contrast to conventional Q-learning algorithms, the DQN employs the target network and experience replay to acquire an unbiased estimation of the mean-square Bellman error (MSBE) utilized in training the Q-network. However, the existing theoretical analysis of DQNs lacks convergence analysis or bypasses the technical challenges by deploying a significantly overparameterized neural network, which is not computationally efficient. This paper provides the first theoretical convergence and sample complexity analysis of the practical setting of DQNs with $\epsilon$-greedy policy. We prove an iterative procedure with decaying $\epsilon$ converges to the optimal Q-value function geometrically. Moreover, a higher level of $\epsilon$ values enlarges the region of convergence but slows down the convergence, while the opposite holds for a lower level of $\epsilon$ values. Experiments justify our established theoretical insights on DQNs.

Many policy optimization approaches in reinforcement learning incorporate a Kullback-Leilbler (KL) divergence to the previous policy, to prevent the policy from changing too quickly. This idea was initially proposed in a seminal paper on Conservative Policy Iteration, with approximations given by algorithms like TRPO and Munchausen Value Iteration (MVI). We continue this line of work by investigating a generalized KL divergence -- called the Tsallis KL divergence -- which use the $q$-logarithm in the definition. The approach is a strict generalization, as $q = 1$ corresponds to the standard KL divergence; $q > 1$ provides a range of new options. We characterize the types of policies learned under the Tsallis KL, and motivate when $q >1$ could be beneficial. To obtain a practical algorithm that incorporates Tsallis KL regularization, we extend MVI, which is one of the simplest approaches to incorporate KL regularization. We show that this generalized MVI($q$) obtains significant improvements over the standard MVI($q = 1$) across 35 Atari games.