Achieving human-level dexterity in robotic grasping remains a challenging endeavor. Robotic hands frequently encounter slippage and deformation during object manipulation, issues rarely encountered by humans due to their sensory receptors, experiential learning, and motor memory. The emulation of the human grasping reflex within robotic hands is referred to as the ``bionic reflex". Past endeavors in the realm of bionic reflex control predominantly relied on model-based and supervised learning approaches, necessitating human intervention during thresholding and labeling tasks. In this study, we introduce an innovative bionic reflex control pipeline, leveraging reinforcement learning (RL); thereby eliminating the need for human intervention during control design. Our proposed bionic reflex controller has been designed and tested on an anthropomorphic hand, manipulating deformable objects in the PyBullet physics simulator, incorporating domain randomization (DR) for enhanced Sim2Real transferability. Our findings underscore the promise of RL as a potent tool for advancing bionic reflex control within anthropomorphic robotic hands. We anticipate that this autonomous, RL-based bionic reflex controller will catalyze the development of dependable and highly efficient robotic and prosthetic hands, revolutionizing human-robot interaction and assistive technologies.

Communication networks able to withstand hostile environments are critically important for disaster relief operations. In this paper, we consider a challenging scenario where drones have been compromised in the supply chain, during their manufacture, and harbour malicious software capable of wide-ranging and infectious disruption. We investigate multi-agent deep reinforcement learning as a tool for learning defensive strategies that maximise communications bandwidth despite continual adversarial interference. Using a public challenge for learning network resilience strategies, we propose a state-of-the-art expert technique and study its superiority over deep reinforcement learning agents. Correspondingly, we identify three specific methods for improving the performance of our learning-based agents: (1) ensuring each observation contains the necessary information, (2) using expert agents to provide a curriculum for learning, and (3) paying close attention to reward. We apply our methods and present a new mixed strategy enabling expert and learning-based agents to work together and improve on all prior results.

Structural pruning has become an integral part of neural network optimization, used to achieve architectural configurations which can be deployed and run more efficiently on embedded devices. Previous results showed that pruning is possible with minimum performance loss by utilizing a reinforcement learning agent which makes decisions about the sparsity level of each neural layer by maximizing as a reward the accuracy of the network. We introduce a novel information-theoretic reward function which minimizes the spatial entropy of convolutional activations. This minimization ultimately acts as a proxy for maintaining accuracy, although these two criteria are not related in any way. Our method shows that there is another possibility to preserve accuracy without the need to directly optimize it in the agent's reward function. In our experiments, we were able to reduce the total number of FLOPS of multiple popular neural network architectures by 5-10x, incurring minimal or no performance drop and being on par with the solution found by maximizing the accuracy.

We consider two-player zero-sum stochastic games and propose a two-timescale $Q$-learning algorithm with function approximation that is payoff-based, convergent, rational, and symmetric between the two players. In two-timescale $Q$-learning, the fast-timescale iterates are updated in spirit to the stochastic gradient descent and the slow-timescale iterates (which we use to compute the policies) are updated by taking a convex combination between its previous iterate and the latest fast-timescale iterate. Introducing the slow timescale as well as its update equation marks as our main algorithmic novelty. In the special case of linear function approximation, we establish, to the best of our knowledge, the first last-iterate finite-sample bound for payoff-based independent learning dynamics of these types. The result implies a polynomial sample complexity to find a Nash equilibrium in such stochastic games. To establish the results, we model our proposed algorithm as a two-timescale stochastic approximation and derive the finite-sample bound through a Lyapunov-based approach. The key novelty lies in constructing a valid Lyapunov function to capture the evolution of the slow-timescale iterates. Specifically, through a change of variable, we show that the update equation of the slow-timescale iterates resembles the classical smoothed best-response dynamics, where the regularized Nash gap serves as a valid Lyapunov function. This insight enables us to construct a valid Lyapunov function via a generalized variant of the Moreau envelope of the regularized Nash gap. The construction of our Lyapunov function might be of broad independent interest in studying the behavior of stochastic approximation algorithms.

Robotic manipulation of deformable linear objects (DLOs) has great potential for applications in diverse fields such as agriculture or industry. However, a major challenge lies in acquiring accurate deformation models that describe the relationship between robot motion and DLO deformations. Such models are difficult to calculate analytically and vary among DLOs. Consequently, manipulating DLOs poses significant challenges, particularly in achieving large deformations that require highly accurate global models. To address these challenges, this paper presents MultiAC6: a new multi Actor-Critic framework for robot action space decomposition to control large 3D deformations of DLOs. In our approach, two deep reinforcement learning (DRL) agents orient and position a robot gripper to deform a DLO into the desired shape. Unlike previous DRL-based studies, MultiAC6 is able to solve the sim-to-real gap, achieving large 3D deformations up to 40 cm in real-world settings. Experimental results also show that MultiAC6 has a 66\% higher success rate than a single-agent approach. Further experimental studies demonstrate that MultiAC6 generalizes well, without retraining, to DLOs with different lengths or materials.

The ability to predict upcoming events has been hypothesized to comprise a key aspect of natural and machine cognition. This is supported by trends in deep reinforcement learning (RL), where self-supervised auxiliary objectives such as prediction are widely used to support representation learning and improve task performance. Here, we study the effects predictive auxiliary objectives have on representation learning across different modules of an RL system and how these mimic representational changes observed in the brain. We find that predictive objectives improve and stabilize learning particularly in resource-limited architectures, and we identify settings where longer predictive horizons better support representational transfer. Furthermore, we find that representational changes in this RL system bear a striking resemblance to changes in neural activity observed in the brain across various experiments. Specifically, we draw a connection between the auxiliary predictive model of the RL system and hippocampus, an area thought to learn a predictive model to support memory-guided behavior. We also connect the encoder network and the value learning network of the RL system to visual cortex and striatum in the brain, respectively. This work demonstrates how representation learning in deep RL systems can provide an interpretable framework for modeling multi-region interactions in the brain. The deep RL perspective taken here also suggests an additional role of the hippocampus in the brain -- that of an auxiliary learning system that benefits representation learning in other regions.

In Reinforcement Learning (RL), enhancing sample efficiency is crucial, particularly in scenarios when data acquisition is costly and risky. In principle, off-policy RL algorithms can improve sample efficiency by allowing multiple updates per environment interaction. However, these multiple updates often lead the model to overfit to earlier interactions, which is referred to as the loss of plasticity. Our study investigates the underlying causes of this phenomenon by dividing plasticity into two aspects. Input plasticity, which denotes the model's adaptability to changing input data, and label plasticity, which denotes the model's adaptability to evolving input-output relationships. Synthetic experiments on the CIFAR-10 dataset reveal that finding smoother minima of loss landscape enhances input plasticity, whereas refined gradient propagation improves label plasticity. Leveraging these findings, we introduce the PLASTIC algorithm, which harmoniously combines techniques to address both concerns. With minimal architectural modifications, PLASTIC achieves competitive performance on benchmarks including Atari-100k and Deepmind Control Suite. This result emphasizes the importance of preserving the model's plasticity to elevate the sample efficiency in RL. The code is available at https://github.com/dojeon-ai/plastic.

Open-endedness stands to benefit from the ability to generate an infinite variety of diverse, challenging environments. One particularly interesting type of challenge is meta-learning ("learning-to-learn"), a hallmark of intelligent behavior. However, the number of meta-learning environments in the literature is limited. Here we describe a parametrized space for simple meta-reinforcement learning (meta-RL) tasks with arbitrary stimuli. The parametrization allows us to randomly generate an arbitrary number of novel simple meta-learning tasks. The parametrization is expressive enough to include many well-known meta-RL tasks, such as bandit problems, the Harlow task, T-mazes, the Daw two-step task and others. Simple extensions allow it to capture tasks based on two-dimensional topological spaces, such as full mazes or find-the-spot domains. We describe a number of randomly generated meta-RL domains of varying complexity and discuss potential issues arising from random generation.

Supervised imitation learning, also known as behavioral cloning, suffers from distribution drift leading to failures during policy execution. One approach to mitigate this issue is to allow an expert to correct the agent's actions during task execution, based on the expert's determination that the agent has reached a `point of no return.' The agent's policy is then retrained using this new corrective data. This approach alone can enable high-performance agents to be learned, but at a substantial cost: the expert must vigilantly observe execution until the policy reaches a specified level of success, and even at that point, there is no guarantee that the policy will always succeed. To address these limitations, we present FIRE (Failure Identification to Reduce Expert Burden in intervention-based learning), a system that can predict when a running policy will fail, halt its execution, and request a correction from the expert. Unlike existing approaches that learn only from expert data, our approach learns from both expert and non-expert data, akin to adversarial learning. We demonstrate experimentally for a series of challenging manipulation tasks that our method is able to recognize state-action pairs that lead to failures. This permits seamless integration into an intervention-based learning system, where we show an order-of-magnitude gain in sample efficiency compared with a state-of-the-art inverse reinforcement learning method and dramatically improved performance over an equivalent amount of data learned with behavioral cloning.

We introduce Onflow, a reinforcement learning technique that enables online optimization of portfolio allocation policies based on gradient flows. We devise dynamic allocations of an investment portfolio to maximize its expected log return while taking into account transaction fees. The portfolio allocation is parameterized through a softmax function, and at each time step, the gradient flow method leads to an ordinary differential equation whose solutions correspond to the updated allocations. This algorithm belongs to the large class of stochastic optimization procedures; we measure its efficiency by comparing our results to the mathematical theoretical values in a log-normal framework and to standard benchmarks from the 'old NYSE' dataset. For log-normal assets, the strategy learned by Onflow, with transaction costs at zero, mimics Markowitz's optimal portfolio and thus the best possible asset allocation strategy. Numerical experiments from the 'old NYSE' dataset show that Onflow leads to dynamic asset allocation strategies whose performances are: a) comparable to benchmark strategies such as Cover's Universal Portfolio or Helmbold et al. "multiplicative updates" approach when transaction costs are zero, and b) better than previous procedures when transaction costs are high. Onflow can even remain efficient in regimes where other dynamical allocation techniques do not work anymore. Therefore, as far as tested, Onflow appears to be a promising dynamic portfolio management strategy based on observed prices only and without any assumption on the laws of distributions of the underlying assets' returns. In particular it could avoid model risk when building a trading strategy.