Vision-based reinforcement learning requires efficient and robust representations of image-based observations, especially when the images contain distracting (task-irrelevant) elements such as shadows, clouds, and light. It becomes more important if those distractions are not exposed during training. We design a Self-Predictive Dynamics (SPD) method to extract task-relevant features efficiently, even in unseen observations after training. SPD uses weak and strong augmentations in parallel, and learns representations by predicting inverse and forward transitions across the two-way augmented versions. In a set of MuJoCo visual control tasks and an autonomous driving task (CARLA), SPD outperforms previous studies in complex observations, and significantly improves the generalization performance for unseen observations. Our code is available at https://github.com/unigary/SPD.

Large reasoning models (LRMs) achieve higher performance on challenging reasoning tasks by generating more tokens at inference time, but this verbosity often wastes computation on easy problems. Existing solutions, including supervised finetuning on shorter traces, user-controlled budgets, or RL with uniform penalties, either require data curation, manual configuration, or treat all problems alike regardless of difficulty. We introduce Adaptive Length Penalty (ALP), a reinforcement learning objective tailoring generation length to per-prompt solve rate. During training, ALP monitors each prompt's online solve rate through multiple rollouts and adds a differentiable penalty whose magnitude scales inversely with that rate, so confident (easy) prompts incur a high cost for extra tokens while hard prompts remain unhindered. Posttraining DeepScaleR-1.5B with ALP cuts average token usage by 50\% without significantly dropping performance. Relative to fixed-budget and uniform penalty baselines, ALP redistributes its reduced budget more intelligently by cutting compute on easy prompts and reallocating saved tokens to difficult ones, delivering higher accuracy on the hardest problems with higher cost.

Robot learning increasingly relies on simulation to advance complex ability such as dexterous manipulations and precise interactions, necessitating high-quality digital assets to bridge the sim-to-real gap. However, existing open-source articulated-object datasets for simulation are limited by insufficient visual realism and low physical fidelity, which hinder their utility for training models mastering robotic tasks in real world. To address these challenges, we introduce ArtVIP, a comprehensive open-source dataset comprising high-quality digital-twin articulated objects, accompanied by indoor-scene assets. Crafted by professional 3D modelers adhering to unified standards, ArtVIP ensures visual realism through precise geometric meshes and high-resolution textures, while physical fidelity is achieved via fine-tuned dynamic parameters. Meanwhile, the dataset pioneers embedded modular interaction behaviors within assets and pixel-level affordance annotations. Feature-map visualization and optical motion capture are employed to quantitatively demonstrate ArtVIP's visual and physical fidelity, with its applicability validated across imitation learning and reinforcement learning experiments. Provided in USD format with detailed production guidelines, ArtVIP is fully open-source, benefiting the research community and advancing robot learning research. Our project is at https://x-humanoid-artvip.github.io/ .

This paper presents a novel reinforcement learning (RL) framework for dynamically optimizing numerical precision in the preconditioned conjugate gradient (CG) method. By modeling precision selection as a Markov Decision Process (MDP), we employ Q-learning to adaptively assign precision levels to key operations, striking an optimal balance between computational efficiency and numerical accuracy, while ensuring stability through double-precision scalar computations and residual computing. In practice, the algorithm is trained on a set of data and subsequently performs inference for precision selection on out-of-sample data, without requiring re-analysis or retraining for new datasets. This enables the method to adapt seamlessly to new problem instances without the computational overhead of recalibration. Our results demonstrate the effectiveness of RL in enhancing solver's performance, marking the first application of RL to mixed-precision numerical methods. The findings highlight the approach's practical advantages, robustness, and scalability, providing valuable insights into its integration with iterative solvers and paving the way for AI-driven advancements in scientific computing.

Large language models (LLMs) excel at many supervised tasks but often struggle with structured reasoning in unfamiliar settings. This discrepancy suggests that standard fine-tuning pipelines may instill narrow, domain-specific heuristics rather than fostering general-purpose thinking strategies. In this work, we propose a "play to learn" framework that fine-tunes LLMs through reinforcement learning on a suite of seven custom logic puzzles, each designed to cultivate distinct reasoning skills such as constraint propagation, spatial consistency, and symbolic deduction. Using a reinforcement learning setup with verifiable rewards, models receive binary feedback based on puzzle correctness, encouraging iterative, hypothesis-driven problem solving. We demonstrate that this training approach significantly improves out-of-distribution performance on a range of mathematical benchmarks, especially for mid-difficulty problems that require multi-step reasoning. Analyses across problem categories and difficulty levels reveal that puzzle training promotes transferable reasoning routines, strengthening algebraic manipulation, geometric inference, and combinatorial logic, while offering limited gains on rote or highly specialized tasks. These findings show that reinforcement learning over logic puzzles reshapes the internal reasoning of LLMs, enabling more robust and compositional generalization without relying on task-specific symbolic tools.

The paper explores the use of Deep Reinforcement Learning (DRL) in stock market trading, focusing on two algorithms: Double Deep Q-Network (DDQN) and Proximal Policy Optimization (PPO) and compares them with Buy and Hold benchmark. It evaluates these algorithms across three currency pairs, the S&P 500 index and Bitcoin, on the daily data in the period of 2019-2023. The results demonstrate DRL's effectiveness in trading and its ability to manage risk by strategically avoiding trades in unfavorable conditions, providing a substantial edge over classical approaches, based on supervised learning in terms of risk-adjusted returns.

Offline reinforcement learning (RL) aims to find optimal policies in dynamic environments in order to maximize the expected total rewards by leveraging pre-collected data. Learning from heterogeneous data is one of the fundamental challenges in offline RL. Traditional methods focus on learning an optimal policy for all individuals with pre-collected data from a single episode or homogeneous batch episodes, and thus, may result in a suboptimal policy for a heterogeneous population. In this paper, we propose an individualized offline policy optimization framework for heterogeneous time-stationary Markov decision processes (MDPs). The proposed heterogeneous model with individual latent variables enables us to efficiently estimate the individual Q-functions, and our Penalized Pessimistic Personalized Policy Learning (P4L) algorithm guarantees a fast rate on the average regret under a weak partial coverage assumption on behavior policies. In addition, our simulation studies and a real data application demonstrate the superior numerical performance of the proposed method compared with existing methods.

Motivated by real-world settings where data collection and policy deployment -- whether for a single agent or across multiple agents -- are costly, we study the problem of on-policy single-agent reinforcement learning (RL) and federated RL (FRL) with a focus on minimizing burn-in costs (the sample sizes needed to reach near-optimal regret) and policy switching or communication costs. In parallel finite-horizon episodic Markov Decision Processes (MDPs) with $S$ states and $A$ actions, existing methods either require superlinear burn-in costs in $S$ and $A$ or fail to achieve logarithmic switching or communication costs. We propose two novel model-free RL algorithms -- Q-EarlySettled-LowCost and FedQ-EarlySettled-LowCost -- that are the first in the literature to simultaneously achieve: (i) the best near-optimal regret among all known model-free RL or FRL algorithms, (ii) low burn-in cost that scales linearly with $S$ and $A$, and (iii) logarithmic policy switching cost for single-agent RL or communication cost for FRL. Additionally, we establish gap-dependent theoretical guarantees for both regret and switching/communication costs, improving or matching the best-known gap-dependent bounds.

Reinforcement learning (RL) has demonstrated remarkable success in enhancing model capabilities, including instruction-following, preference learning, and reasoning. Yet despite its empirical successes, the mechanisms by which RL improves reasoning abilities remain poorly understood. We present a systematic study of Reinforcement Learning with Verifiable Rewards (RLVR), showing that its primary benefit comes from optimizing the selection of existing reasoning patterns. Through extensive experiments, we demonstrate that RLVR-trained models preferentially adopt high-success-rate reasoning patterns while mostly maintaining stable performance on individual patterns. We further develop theoretical analyses on the convergence and training dynamics of RLVR based on a simplified question-reason-answer model. We study the gradient flow and show that RLVR can indeed find the solution that selects the reason pattern with the highest success rate. Besides, our theoretical results reveal two distinct regimes regarding the convergence of RLVR training: (1) rapid convergence for models with relatively strong initial reasoning capabilities versus (2) slower optimization dynamics for weaker models. Furthermore, we show that the slower optimization for weaker models can be mitigated by applying the supervised fine-tuning (SFT) before RLVR, when using a feasibly high-quality SFT dataset. We validate the theoretical findings through extensive experiments. This work advances our theoretical understanding of RL's role in LLM fine-tuning and offers insights for further enhancing reasoning capabilities.

Maintaining an active lifestyle is vital for quality of life, yet challenging for wheelchair users. For instance, powered wheelchairs face increasing risks of obesity and deconditioning due to inactivity. Conversely, manual wheelchair users, who propel the wheelchair by pushing the wheelchair's handrims, often face upper extremity injuries from repetitive motions. These challenges underscore the need for a mobility system that promotes activity while minimizing injury risk. Maintaining optimal exertion during wheelchair use enhances health benefits and engagement, yet the variations in individual physiological responses complicate exertion optimization. To address this, we introduce PulseRide, a novel wheelchair system that provides personalized assistance based on each user's physiological responses, helping them maintain their physical exertion goals. Unlike conventional assistive systems focused on obstacle avoidance and navigation, PulseRide integrates real-time physiological data-such as heart rate and ECG-with wheelchair speed to deliver adaptive assistance. Using a human-in-the-loop reinforcement learning approach with Deep Q-Network algorithm (DQN), the system adjusts push assistance to keep users within a moderate activity range without under- or over-exertion. We conducted preliminary tests with 10 users on various terrains, including carpet and slate, to assess PulseRide's effectiveness. Our findings show that, for individual users, PulseRide maintains heart rates within the moderate activity zone as much as 71.7 percent longer than manual wheelchairs. Among all users, we observed an average reduction in muscle contractions of 41.86 percent, delaying fatigue onset and enhancing overall comfort and engagement. These results indicate that PulseRide offers a healthier, adaptive mobility solution, bridging the gap between passive and physically taxing mobility options.