Yu Wang

Tonghe Zhang, Chao Yu, Sichang Su et al.

We propose ReinFlow, a simple yet effective online reinforcement learning (RL) framework that fine-tunes a family of flow matching policies for continuous robotic control. Derived from rigorous RL theory, ReinFlow injects learnable noise into a flow policy's deterministic path, converting the flow into a discrete-time Markov Process for exact and straightforward likelihood computation. This conversion facilitates exploration and ensures training stability, enabling ReinFlow to fine-tune diverse flow model variants, including Rectified Flow [35] and Shortcut Models [19], particularly at very few or even one denoising step. We benchmark ReinFlow in representative locomotion and manipulation tasks, including long-horizon planning with visual input and sparse reward. The episode reward of Rectified Flow policies obtained an average net growth of 135.36% after fine-tuning in challenging legged locomotion tasks while saving denoising steps and 82.63% of wall time compared to state-of-the-art diffusion RL fine-tuning method DPPO [43]. The success rate of the Shortcut Model policies in state and visual manipulation tasks achieved an average net increase of 40.34% after fine-tuning with ReinFlow at four or even one denoising step, whose performance is comparable to fine-tuned DDIM policies while saving computation time for an average of 23.20%. Project webpage: https://reinflow.github.io/

Jiayu Chen, Guosheng Li, Chao Yu et al.

This paper addresses multi-UAV pursuit-evasion, where a group of drones cooperates to capture a fast evader in a confined environment with obstacles. Existing heuristic algorithms, which simplify the pursuit-evasion problem, often lack expressive coordination strategies and struggle to capture the evader in extreme scenarios, such as when the evader moves at high speeds. In contrast, reinforcement learning (RL) has been applied to this problem and has the potential to obtain highly cooperative capture strategies. However, RL-based methods face challenges in training for complex 3-dimensional scenarios with diverse task settings due to the vast exploration space. The dynamics constraints of drones further restrict the ability of reinforcement learning to acquire high-performance capture strategies. In this work, we introduce a dual curriculum learning framework, named DualCL, which addresses multi-UAV pursuit-evasion in diverse environments and demonstrates zero-shot transfer ability to unseen scenarios. DualCL comprises two main components: the Intrinsic Parameter Curriculum Proposer, which progressively suggests intrinsic parameters from easy to hard to improve the capture capability of drones, and the External Environment Generator, tasked with exploring unresolved scenarios and generating appropriate training distributions of external environment parameters. The simulation experimental results show that DualCL significantly outperforms baseline methods, achieving over 90% capture rate and reducing the capture timestep by at least 27.5% in the training scenarios. Additionally, it exhibits the best zero-shot generalization ability in unseen environments. Moreover, we demonstrate the transferability of our pursuit strategy from simulation to real-world environments. Further details can be found on the project website at https://sites.google.com/view/dualcl.

Yusheng Liao, Shuyang Jiang, Zhe Chen et al.

Large language models (LLMs) have shown substantial progress in natural language understanding and generation, proving valuable especially in the medical field. Despite advancements, challenges persist due to the complexity and diversity inherent in medical tasks, which can be categorized as knowledge-intensive tasks and alignment-required tasks. Previous approaches either ignore the latter task or focus on a minority of tasks and hence lose generalization. To address these drawbacks, we propose a progressive fine-tuning pipeline. This pipeline employs a Knowledge Aggregator and a Noise aggregator to encode diverse knowledge in the first stage and filter out detrimental information. In the second stage, we drop the Noise Aggregator to avoid the interference of suboptimal representation and leverage an additional alignment module optimized towards an orthogonal direction to the knowledge space to mitigate knowledge forgetting. Based on this two-stage paradigm, we proposed a Medical LLM through decoupling Clinical Alignment and Knowledge Aggregation (MedCare), which is designed to achieve state-of-the-art (SOTA) performance on over 20 medical tasks, as well as SOTA results on specific medical alignment tasks. Various model sizes of MedCare (1.8B, 7B, 14B) all demonstrate significant improvements over existing models with similar model sizes.

Yu Wang, Shu Jiang, Weiman Lin et al.

This paper presents the design of a tune-free (human-out-of-the-loop parameter tuning) control framework, aiming at accelerating large scale autonomous driving system deployed on various vehicles and driving environments. The framework consists of three machine-learning-based procedures, which jointly automate the control parameter tuning for autonomous driving, including: a learning-based dynamic modeling procedure, to enable the control-in-the-loop simulation with highly accurate vehicle dynamics for parameter tuning; a learning-based open-loop mapping procedure, to solve the feedforward control parameters tuning; and more significantly, a Bayesian-optimization-based closed-loop parameter tuning procedure, to automatically tune feedback control (PID, LQR, MRAC, MPC, etc.) parameters in simulation environment. The paper shows an improvement in control performance with a significant increase in parameter tuning efficiency, in both simulation and road tests. This framework has been validated on different vehicles in US and China.

Kumpati S. Narendra, Yu Wang, Wei Chen

Recently, a new approach to the adaptive control of linear time-invariant plants with unknown parameters (referred to as second level adaptation), was introduced by Han and Narendra in [1]. Based on $N (\geq m+1)$ fixed or adaptive models of the plant, where $m$ is the dimension of the unknown parameter vector, an unknown parameter vector $α\in R^{N}$ is estimated in the new approach, and in turn, is used to control the overall system. Simulation studies were presented in [1] to demonstrate that the new method is significantly better than those that are currently in use. In this paper, we undertake a more detailed examination of the theoretical and practical advantages claimed for the new method. In particular, the need for many models, the proof of stability, and the improvement in performance and robustness are explored in depth both theoretically and experimentally.