This study addresses the challenge of forming effective groups in collaborative problem-solving environments. Recognizing the complexity of human interactions and the necessity for efficient collaboration, we propose a novel approach leveraging graph theory and reinforcement learning. Our methodology involves constructing a graph from a dataset where nodes represent participants, and edges signify the interactions between them. We conceptualize each participant as an agent within a reinforcement learning framework, aiming to learn an optimal graph structure that reflects effective group dynamics. Clustering techniques are employed to delineate clear group structures based on the learned graph. Our approach provides theoretical solutions based on evaluation metrics and graph measurements, offering insights into potential improvements in group effectiveness and reductions in conflict incidences. This research contributes to the fields of collaborative work and educational psychology by presenting a data-driven, analytical approach to group formation. It has practical implications for organizational team building, classroom settings, and any collaborative scenario where group dynamics are crucial. The study opens new avenues for exploring the application of graph theory and reinforcement learning in social and behavioral sciences, highlighting the potential for empirical validation in future work.

In preference-based reinforcement learning (PbRL), a reward function is learned from a type of human feedback called preference. To expedite preference collection, recent works have leveraged \emph{offline preferences}, which are preferences collected for some offline data. In this scenario, the learned reward function is fitted on the offline data. If a learning agent exhibits behaviors that do not overlap with the offline data, the learned reward function may encounter generalizability issues. To address this problem, the present study introduces a framework that consolidates offline preferences and \emph{virtual preferences} for PbRL, which are comparisons between the agent's behaviors and the offline data. Critically, the reward function can track the agent's behaviors using the virtual preferences, thereby offering well-aligned guidance to the agent. Through experiments on continuous control tasks, this study demonstrates the effectiveness of incorporating the virtual preferences in PbRL.

Crowd navigation has received significant research attention in recent years, especially DRL-based methods. While single-robot crowd scenarios have dominated research, they offer limited applicability to real-world complexities. The heterogeneity of interaction among multiple agent categories, like in decentralized multi-robot pedestrian scenarios, are frequently disregarded. This "interaction blind spot" hinders generalizability and restricts progress towards robust navigation algorithms. In this paper, we propose a heterogeneous relational deep reinforcement learning(HeR-DRL), based on customised heterogeneous GNN, in order to improve navigation strategies in decentralized multi-robot crowd navigation. Firstly, we devised a method for constructing robot-crowd heterogenous relation graph that effectively simulates the heterogeneous pair-wise interaction relationships. We proposed a new heterogeneous graph neural network for transferring and aggregating the heterogeneous state information. Finally, we incorporate the encoded information into deep reinforcement learning to explore the optimal policy. HeR-DRL are rigorously evaluated through comparing it to state-of-the-art algorithms in both single-robot and multi-robot circle crowssing scenario. The experimental results demonstrate that HeR-DRL surpasses the state-of-the-art approaches in overall performance, particularly excelling in safety and comfort metrics. This underscores the significance of interaction heterogeneity for crowd navigation. The source code will be publicly released in https://github.com/Zhouxy-Debugging-Den/HeR-DRL.

It is common for us to feel pressure in a competition environment, which arises from the desire to obtain success comparing with other individuals or opponents. Although we might get anxious under the pressure, it could also be a drive for us to stimulate our potentials to the best in order to keep up with others. Inspired by this, we propose a competitive learning framework which is able to help individual robot to acquire knowledge from the competition, fully stimulating its dynamics potential in the race. Specifically, the competition information among competitors is introduced as the additional auxiliary signal to learn advantaged actions. We further build a Multiagent-Race environment, and extensive experiments are conducted, demonstrating that robots trained in competitive environments outperform ones that are trained with SoTA algorithms in single robot environment.

Deformable objects present several challenges to the field of robotic manipulation. One of the tasks that best encapsulates the difficulties arising due to non-rigid behavior is shape control, which requires driving an object to a desired shape. While shape-servoing methods have been shown successful in contexts with approximately linear behavior, they can fail in tasks with more complex dynamics. We investigate an alternative approach, using offline RL to solve a planar shape control problem of a Deformable Linear Object (DLO). To evaluate the effect of material properties, two DLOs are tested namely a soft rope and an elastic cord. We frame this task as a goal-conditioned offline RL problem, and aim to learn to generalize to unseen goal shapes. Data collection and augmentation procedures are proposed to limit the amount of experimental data which needs to be collected with the real robot. We evaluate the amount of augmentation needed to achieve the best results, and test the effect of regularization through behavior cloning on the TD3+BC algorithm. Finally, we show that the proposed approach is able to outperform a shape-servoing baseline in a curvature inversion experiment.

Text-based games (TBGs) have emerged as an important collection of NLP tasks, requiring reinforcement learning (RL) agents to combine natural language understanding with reasoning. A key challenge for agents attempting to solve such tasks is to generalize across multiple games and demonstrate good performance on both seen and unseen objects. Purely deep-RL-based approaches may perform well on seen objects; however, they fail to showcase the same performance on unseen objects. Commonsense-infused deep-RL agents may work better on unseen data; unfortunately, their policies are often not interpretable or easily transferable. To tackle these issues, in this paper, we present EXPLORER which is an exploration-guided reasoning agent for textual reinforcement learning. EXPLORER is neurosymbolic in nature, as it relies on a neural module for exploration and a symbolic module for exploitation. It can also learn generalized symbolic policies and perform well over unseen data. Our experiments show that EXPLORER outperforms the baseline agents on Text-World cooking (TW-Cooking) and Text-World Commonsense (TWC) games.

Nonprehensile manipulation is essential for manipulating objects that are too thin, large, or otherwise ungraspable in the wild. To sidestep the difficulty of contact modeling in conventional modeling-based approaches, reinforcement learning (RL) has recently emerged as a promising alternative. However, previous RL approaches either lack the ability to generalize over diverse object shapes, or use simple action primitives that limit the diversity of robot motions. Furthermore, using RL over diverse object geometry is challenging due to the high cost of training a policy that takes in high-dimensional sensory inputs. We propose a novel contact-based object representation and pretraining pipeline to tackle this. To enable massively parallel training, we leverage a lightweight patch-based transformer architecture for our encoder that processes point clouds, thus scaling our training across thousands of environments. Compared to learning from scratch, or other shape representation baselines, our representation facilitates both time-and data-efficient learning. We validate the efficacy of our overall system by zero-shot transferring the trained policy to novel real-world objects. Robot manipulators can transform our day-to-day lives by alleviating humans from taxing physical labor. Despite this potential, they have remained largely confined to picking and placing objects in limited settings such as manufacturing or logistics facilities.

Reinforcement Learning- (RL-)based motion planning has recently shown the potential to outperform traditional approaches from autonomous navigation to robot manipulation. In this work, we focus on a motion planning task for an evasive target in a partially observable multi-agent adversarial pursuit-evasion games (PEG). These pursuit-evasion problems are relevant to various applications, such as search and rescue operations and surveillance robots, where robots must effectively plan their actions to gather intelligence or accomplish mission tasks while avoiding detection or capture themselves. We propose a hierarchical architecture that integrates a high-level diffusion model to plan global paths responsive to environment data while a low-level RL algorithm reasons about evasive versus global path-following behavior. Our approach outperforms baselines by 51.2% by leveraging the diffusion model to guide the RL algorithm for more efficient exploration and improves the explanability and predictability.

Reinforcement Learning from Human Feedback (RLHF) has proven to be a strong method to align Pretrained Large Language Models (LLMs) with human preferences. But training models with RLHF is computationally expensive, and an overall complex process. In this work, we study RLHF where the underlying models are trained using the parameter efficient method of Low-Rank Adaptation (LoRA) introduced by Hu et al. [2021]. We investigate the setup of "Parameter Efficient Reinforcement Learning" (PERL), in which we perform reward model training and reinforcement learning using LoRA. We compare PERL to conventional fine-tuning (full-tuning) across various configurations for 7 benchmarks, including 2 novel datasets, of reward modeling and reinforcement learning. We find that PERL performs on par with the conventional RLHF setting, while training faster, and with less memory. This enables the high performance of RLHF, while reducing the computational burden that limits its adoption as an alignment technique for Large Language Models. We also release 2 novel thumbs up/down preference datasets: "Taskmaster Coffee", and "Taskmaster Ticketing" to promote research around RLHF.

Diabetes profoundly affects human life and health, regardless of country, age, or gender, and is one of the leading causes of death and disability worldwide [1]. From 1990 to 2021, the age-standardized prevalence of diabetes increased by 90.5 % globally, with increases of more than 100 % in several regions, and it is projected that by 2050, there will be 1.31 billion people with diabetes worldwide [1]. Furthermore, people with diabetes have more than twice the normal risk of early death, resulting in an estimated 150-500 million deaths around the world each year, while generating approximately 12% of health expenditure ($966 billion) [2, 3]. The rising prevalence and serious health and economic hazards have attracted the attention of scientists around the globe, and as a result, the number of studies on diabetes is increasing. The pancreas of a diabetic does not produce or produces very little insulin, or the insulin produced is not used efficiently, leading to high BG and a variety of life-threatening complications such as cardiovascular disease, nerve damage, kidney damage, lower limb amputations, and eye disease leading to decreased vision and even blindness [3]. BG control is their basic treatment, as well as the basis for preventing and treating diabetic complications. Patients mainly maintain the stability of BG by injecting insulin. However, this traditional self-management is usually cumbersome and challenging, as it requires patients to measure their BG levels several times a day, while they suffer from many of the aforementioned complications [2].