On manifolds with metrics, minimizing geodesics, or shortest paths, are minimum-length curves connecting points. Various real-world tasks can be reduced to the generation of geodesics on manifolds. Examples include time-optimal path planning on sloping ground [Matsumoto, 1989], robot motion planning under various constraints [LaValle, 2006, Ratliff et al., 2015], physical systems [Pfeifer, 2019], the Wasserstein distance [Agueh, 2012], and image morphing [Michelis and Becker, 2021, Effland et al., 2021]. Typically, metrics are only known infinitesimally (a form of a Riemannian or Finsler metric), and their distance functions are not known beforehand. Computation of geodesics by solving optimization problems or differential equations is generally computationally costly and requires an explicit form of the metric, or at least, values of its differentials.

In-hand tool manipulation is an operation that not only manipulates a tool within the hand (i.e., in-hand manipulation) but also achieves a grasp suitable for a task after the manipulation. This study aims to achieve an in-hand tool manipulation skill through deep reinforcement learning. The difficulty of learning the skill arises because this manipulation requires (A) exploring long-term contact-state changes to achieve the desired grasp and (B) highly-varied motions depending on the contact-state transition. (A) leads to a sparsity of a reward on a successful grasp, and (B) requires an RL agent to explore widely within the state-action space to learn highly-varied actions, leading to sample inefficiency. To address these issues, this study proposes Action Primitives based on Contact-state Transition (APriCoT). APriCoT decomposes the manipulation into short-term action primitives by describing the operation as a contact-state transition based on three action representations (detach, crossover, attach). In each action primitive, fingers are required to perform short-term and similar actions. By training a policy for each primitive, we can mitigate the issues from (A) and (B). This study focuses on a fundamental operation as an example of in-hand tool manipulation: rotating an elongated object grasped with a precision grasp by half a turn to achieve the initial grasp. Experimental results demonstrated that ours succeeded in both the rotation and the achievement of the desired grasp, unlike existing studies. Additionally, it was found that the policy was robust to changes in object shape.

Text-to-image generative models have recently attracted considerable interest, enabling the synthesis of high-quality images from textual prompts. However, these models often lack the capability to generate specific subjects from given reference images or to synthesize novel renditions under varying conditions. Methods like DreamBooth and Subject-driven Text-to-Image (SuTI) have made significant progress in this area. Yet, both approaches primarily focus on enhancing similarity to reference images and require expensive setups, often overlooking the need for efficient training and avoiding overfitting to the reference images. In this work, we present the $\lambda$-Harmonic reward function, which provides a reliable reward signal and enables early stopping for faster training and effective regularization. By combining the Bradley-Terry preference model, the $\lambda$-Harmonic reward function also provides preference labels for subject-driven generation tasks. We propose Reward Preference Optimization (RPO), which offers a simpler setup (requiring only $3\%$ of the negative samples used by DreamBooth) and fewer gradient steps for fine-tuning. Unlike most existing methods, our approach does not require training a text encoder or optimizing text embeddings and achieves text-image alignment by fine-tuning only the U-Net component. Empirically, $\lambda$-Harmonic proves to be a reliable approach for model selection in subject-driven generation tasks. Based on preference labels and early stopping validation from the $\lambda$-Harmonic reward function, our algorithm achieves a state-of-the-art CLIP-I score of 0.833 and a CLIP-T score of 0.314 on DreamBench.

The successor state measure is a central object of study in reinforcement learning (RL). A common statement of the objective is to find the policy that induces the state occupancy measure with the highest expected reward [4, 3, 6, 5, 7]. The state occupancy measure (SOM) has also received considerable attention in the RL theory community, as a number of provably efficient exploration schemes revolve around regularizing the state occupancy measure [1, 2, 8]. We explore a closely related concept, the state successor measure (SSM), which is the probability distribution over future states, given that the agent is currently at state s and takes action a. Despite their utility, the problem of learning the successor measure or state occupancy measure has received relatively little attention in the empirical RL community. While the full reasons for this are difficult to pin down, we argue that it is in large part due to the lack of an expressive and learnable representation that can be easily normalized. We argue that diffusion models can address this deficiency.

Modern reinforcement learning has been conditioned by at least three dogmas. The first is the environment spotlight, which refers to our tendency to focus on modeling environments rather than agents. The second is our treatment of learning as finding the solution to a task, rather than adaptation. The third is the reward hypothesis, which states that all goals and purposes can be well thought of as maximization of a reward signal. These three dogmas shape much of what we think of as the science of reinforcement learning. While each of the dogmas have played an important role in developing the field, it is time we bring them to the surface and reflect on whether they belong as basic ingredients of our scientific paradigm. In order to realize the potential of reinforcement learning as a canonical frame for researching intelligent agents, we suggest that it is time we shed dogmas one and two entirely, and embrace a nuanced approach to the third.

With the increasing demand for multiple applications on internet of vehicles. It requires vehicles to carry out multiple computing tasks in real time. However, due to the insufficient computing capability of vehicles themselves, offloading tasks to vehicular edge computing (VEC) servers and allocating computing resources to tasks becomes a challenge. In this paper, a multi task digital twin (DT) VEC network is established. By using DT to develop offloading strategies and resource allocation strategies for multiple tasks of each vehicle in a single slot, an optimization problem is constructed. To solve it, we propose a multi-agent reinforcement learning method on the task offloading and resource allocation. Numerous experiments demonstrate that our method is effective compared to other benchmark algorithms.

Constrained Reinforcement Learning (CRL) tackles sequential decision-making problems where agents are required to achieve goals by maximizing the expected return while meeting domain-specific constraints, which are often formulated as expected costs. In this setting, policy-based methods are widely used since they come with several advantages when dealing with continuous-control problems. These methods search in the policy space with an action-based or parameter-based exploration strategy, depending on whether they learn directly the parameters of a stochastic policy or those of a stochastic hyperpolicy. In this paper, we propose a general framework for addressing CRL problems via gradient-based primal-dual algorithms, relying on an alternate ascent/descent scheme with dual-variable regularization. We introduce an exploration-agnostic algorithm, called C-PG, which exhibits global last-iterate convergence guarantees under (weak) gradient domination assumptions, improving and generalizing existing results. Then, we design C-PGAE and C-PGPE, the action-based and the parameter-based versions of C-PG, respectively, and we illustrate how they naturally extend to constraints defined in terms of risk measures over the costs, as it is often requested in safety-critical scenarios. Finally, we numerically validate our algorithms on constrained control problems, and compare them with state-of-the-art baselines, demonstrating their effectiveness.

The Value Iteration (VI) algorithm is an iterative procedure to compute the value function of a Markov decision process, and is the basis of many reinforcement learning (RL) algorithms as well. As the error convergence rate of VI as a function of iteration $k$ is $O(\gamma^k)$, it is slow when the discount factor $\gamma$ is close to $1$. To accelerate the computation of the value function, we propose Deflated Dynamics Value Iteration (DDVI). DDVI uses matrix splitting and matrix deflation techniques to effectively remove (deflate) the top $s$ dominant eigen-structure of the transition matrix $\mathcal{P}^{\pi}$. We prove that this leads to a $\tilde{O}(\gamma^k |\lambda_{s+1}|^k)$ convergence rate, where $\lambda_{s+1}$is $(s+1)$-th largest eigenvalue of the dynamics matrix. We then extend DDVI to the RL setting and present Deflated Dynamics Temporal Difference (DDTD) algorithm. We empirically show the effectiveness of the proposed algorithms.

Physically-simulated models for human motion can generate high-quality responsive character animations, often in real-time. Natural language serves as a flexible interface for controlling these models, allowing expert and non-expert users to quickly create and edit their animations. Many recent physics-based animation methods, including those that use text interfaces, train control policies using reinforcement learning (RL). However, scaling these methods beyond several hundred motions has remained challenging. Meanwhile, kinematic animation models are able to successfully learn from thousands of diverse motions by leveraging supervised learning methods. Inspired by these successes, in this work we introduce SuperPADL, a scalable framework for physics-based text-to-motion that leverages both RL and supervised learning to train controllers on thousands of diverse motion clips. SuperPADL is trained in stages using progressive distillation, starting with a large number of specialized experts using RL. These experts are then iteratively distilled into larger, more robust policies using a combination of reinforcement learning and supervised learning. Our final SuperPADL controller is trained on a dataset containing over 5000 skills and runs in real time on a consumer GPU. Moreover, our policy can naturally transition between skills, allowing for users to interactively craft multi-stage animations. We experimentally demonstrate that SuperPADL significantly outperforms RL-based baselines at this large data scale.

Currently, traffic signal control (TSC) methods based on reinforcement learning (RL) have proven superior to traditional methods. However, most RL methods face difficulties when applied in the real world due to three factors: input, output, and the cycle-flow relation. The industry's observable input is much more limited than simulation-based RL methods. For real-world solutions, only flow can be reliably collected, whereas common RL methods need more. For the output action, most RL methods focus on acyclic control, which real-world signal controllers do not support. Most importantly, industry standards require a consistent cycle-flow relationship: non-decreasing and different response strategies for low, medium, and high-level flows, which is ignored by the RL methods. To narrow the gap between RL methods and industry standards, we innovatively propose to use industry solutions to guide the RL agent. Specifically, we design behavior cloning and curriculum learning to guide the agent to mimic and meet industry requirements and, at the same time, leverage the power of exploration and exploitation in RL for better performance. We theoretically prove that such guidance can largely decrease the sample complexity to polynomials in the horizon when searching for an optimal policy. Our rigid experiments show that our method has good cycle-flow relation and superior performance.