DORA: Object Affordance-Guided Reinforcement Learning for Dexterous Robotic Manipulation

Dexterous robotic manipulation remains a longstanding challenge in robotics due to the high dimensionality of control spaces and the semantic complexity of object interaction. In this paper, we propose an object affordance-guided reinforcement learning framework that enables a multi-fingered robotic hand to learn human-like manipulation strategies more efficiently. By leveraging object affordance maps, our approach generates semantically meaningful grasp pose candidates that serve as both policy constraints and priors during training. We introduce a voting-based grasp classification mechanism to ensure functional alignment between grasp configurations and object affordance regions. Furthermore, we incorporate these constraints into a generalizable RL pipeline and design a reward function that unifies affordance-awareness with task-specific objectives. Experimental results across three manipulation tasks - cube grasping, jug grasping and lifting, and hammer use - demonstrate that our affordance-guided approach improves task success rates by an average of 15.4% compared to baselines. These findings highlight the critical role of object affordance priors in enhancing sample efficiency and learning generalizable, semantically grounded manipulation policies. For more details, please visit our project website https://sites.google.com/view/dora-manip.

Qwen3 Technical Report

In this work, we present Qwen3, the latest version of the Qwen model family. Qwen3 comprises a series of large language models (LLMs) designed to advance performance, efficiency, and multilingual capabilities. The Qwen3 series includes models of both dense and Mixture-of-Expert (MoE) architectures, with parameter scales ranging from 0.6 to 235 billion. A key innovation in Qwen3 is the integration of thinking mode (for complex, multi-step reasoning) and non-thinking mode (for rapid, context-driven responses) into a unified framework. This eliminates the need to switch between different models--such as chat-optimized models (e.g., GPT-4o) and dedicated reasoning models (e.g., QwQ-32B)--and enables dynamic mode switching based on user queries or chat templates. Meanwhile, Qwen3 introduces a thinking budget mechanism, allowing users to allocate computational resources adaptively during inference, thereby balancing latency and performance based on task complexity. Moreover, by leveraging the knowledge from the flagship models, we significantly reduce the computational resources required to build smaller-scale models, while ensuring their highly competitive performance. Empirical evaluations demonstrate that Qwen3 achieves state-of-the-art results across diverse benchmarks, including tasks in code generation, mathematical reasoning, agent tasks, etc., competitive against larger MoE models and proprietary models. Compared to its predecessor Qwen2.5, Qwen3 expands multilingual support from 29 to 119 languages and dialects, enhancing global accessibility through improved cross-lingual understanding and generation capabilities. To facilitate reproducibility and community-driven research and development, all Qwen3 models are publicly accessible under Apache 2.0.

Multi-segment Soft Robot Control via Deep Koopman-based Model Predictive Control

Soft robots, compared to regular rigid robots, as their multiple segments with soft materials bring flexibility and compliance, have the advantages of safe interaction and dexterous operation in the environment. However, due to its characteristics of high dimensional, nonlinearity, time-varying nature, and infinite degree of freedom, it has been challenges in achieving precise and dynamic control such as trajectory tracking and position reaching. To address these challenges, we propose a framework of Deep Koopman-based Model Predictive Control (DK-MPC) for handling multi-segment soft robots. We first employ a deep learning approach with sampling data to approximate the Koopman operator, which therefore linearizes the high-dimensional nonlinear dynamics of the soft robots into a finite-dimensional linear representation. Secondly, this linearized model is utilized within a model predictive control framework to compute optimal control inputs that minimize the tracking error between the desired and actual state trajectories. The real-world experiments on the soft robot "Chordata" demonstrate that DK-MPC could achieve high-precision control, showing the potential of DK-MPC for future applications to soft robots.

A Survey of Robotic Navigation and Manipulation with Physics Simulators in the Era of Embodied AI

Navigation and manipulation are core capabilities in Embodied AI, yet training agents with these capabilities in the real world faces high costs and time complexity. Therefore, sim-to-real transfer has emerged as a key approach, yet the sim-to-real gap persists. This survey examines how physics simulators address this gap by analyzing their properties overlooked in previous surveys. We also analyze their features for navigation and manipulation tasks, along with hardware requirements. Additionally, we offer a resource with benchmark datasets, metrics, simulation platforms, and cutting-edge methods-such as world models and geometric equivariance-to help researchers select suitable tools while accounting for hardware constraints.

SoccerDiffusion: Toward Learning End-to-End Humanoid Robot Soccer from Gameplay Recordings

This paper introduces SoccerDiffusion, a transformer-based diffusion model designed to learn end-to-end control policies for humanoid robot soccer directly from real-world gameplay recordings. Using data collected from RoboCup competitions, the model predicts joint command trajectories from multi-modal sensor inputs, including vision, proprioception, and game state. We employ a distillation technique to enable real-time inference on embedded platforms that reduces the multi-step diffusion process to a single step. Our results demonstrate the model's ability to replicate complex motion behaviors such as walking, kicking, and fall recovery both in simulation and on physical robots. Although high-level tactical behavior remains limited, this work provides a robust foundation for subsequent reinforcement learning or preference optimization methods. We release the dataset, pretrained models, and code under: https://bit-bots.github.io/SoccerDiffusion

FABG : End-to-end Imitation Learning for Embodied Affective Human-Robot Interaction

This paper proposes FABG (Facial Affective Behavior Generation), an end-to-end imitation learning system for human-robot interaction, designed to generate natural and fluid facial affective behaviors. In interaction, effectively obtaining high-quality demonstrations remains a challenge. In this work, we develop an immersive virtual reality (VR) demonstration system that allows operators to perceive stereoscopic environments. This system ensures "the operator's visual perception matches the robot's sensory input" and "the operator's actions directly determine the robot's behaviors" - as if the operator replaces the robot in human interaction engagements. We propose a prediction-driven latency compensation strategy to reduce robotic reaction delays and enhance interaction fluency. FABG naturally acquires human interactive behaviors and subconscious motions driven by intuition, eliminating manual behavior scripting. We deploy FABG on a real-world 25-degree-of-freedom (DoF) humanoid robot, validating its effectiveness through four fundamental interaction tasks: expression response, dynamic gaze, foveated attention, and gesture recognition, supported by data collection and policy training. Project website: https://cybergenies.github.io

Catching Spinning Table Tennis Balls in Simulation with End-to-End Curriculum Reinforcement Learning

The game of table tennis is renowned for its extremely high spin rate, but most table tennis robots today struggle to handle balls with such rapid spin. To address this issue, we have contributed a series of methods, including: 1. Curriculum Reinforcement Learning (RL): This method helps the table tennis robot learn to play table tennis progressively from easy to difficult tasks. 2. Analysis of Spinning Table Tennis Ball Collisions: We have conducted a physics-based analysis to generate more realistic trajectories of spinning table tennis balls after collision. 3. Definition of Trajectory States: The definition of trajectory states aids in setting up the reward function. 4. Selection of Valid Rally Trajectories: We have introduced a valid rally trajectory selection scheme to ensure that the robot's training is not influenced by abnormal trajectories. 5. Reality-to-Simulation (Real2Sim) Transfer: This scheme is employed to validate the trained robot's ability to handle spinning balls in real-world scenarios. With Real2Sim, the deployment costs for robotic reinforcement learning can be further reduced. Moreover, the trajectory-state-based reward function is not limited to table tennis robots; it can be generalized to a wide range of cyclical tasks. To validate our robot's ability to handle spinning balls, the Real2Sim experiments were conducted. For the specific video link of the experiment, please refer to the supplementary materials.

RoboBERT: An End-to-end Multimodal Robotic Manipulation Model

Embodied intelligence integrates multiple modalities, enabling agents to understand images, language, and actions simultaneously. However, existing models always depend on additional datasets or extensive pre-training to maximize performance improvements, consuming abundant training time and expensive hardware cost. To tackle this issue, we present RoboBERT, a novel end-to-end robotic manipulation model integrated with a unique training strategy. This model utilizes a CNN-based diffusion policy, enhancing and stabilizing the effectiveness of this model by separating training processes for different modalities. It also underscores the importance of data augmentation, verifying various techniques to significantly boost performance. Unlike models that depend on extra data or large foundation models, RoboBERT achieves a highly competitive success rate while using only language-labeled expert demonstrations and maintaining a relatively smaller model size. Specifically, RoboBERT achieves an average length of 4.52 on the CALVIN benchmark for \(ABCD \rightarrow D\) task, setting a new state-of-the-art (SOTA) record. Furthermore, when tested on a real robot, the model demonstrates superior performance, achieving a higher success rate than other methods trained with the same data. We propose that these concepts and methodologies of RoboBERT demonstrate extensive versatility and compatibility, contributing significantly to the development of lightweight multimodal robotic models. The code can be accessed on https://github.com/PeterWangsicheng/RoboBERT

Qwen2.5-1M Technical Report

We introduce Qwen2.5-1M, a series of models that extend the context length to 1 million tokens. Compared to the previous 128K version, the Qwen2.5-1M series have significantly enhanced long-context capabilities through long-context pre-training and post-training. Key techniques such as long data synthesis, progressive pre-training, and multi-stage supervised fine-tuning are employed to effectively enhance long-context performance while reducing training costs. To promote the use of long-context models among a broader user base, we present and open-source our inference framework. This framework includes a length extrapolation method that can expand the model context lengths by at least four times, or even more, without additional training. To reduce inference costs, we implement a sparse attention method along with chunked prefill optimization for deployment scenarios and a sparsity refinement method to improve precision. Additionally, we detail our optimizations in the inference engine, including kernel optimization, pipeline parallelism, and scheduling optimization, which significantly enhance overall inference performance. By leveraging our inference framework, the Qwen2.5-1M models achieve a remarkable 3x to 7x prefill speedup in scenarios with 1 million tokens of context. This framework provides an efficient and powerful solution for developing applications that require long-context processing using open-source models. The Qwen2.5-1M series currently includes the open-source models Qwen2.5-7B-Instruct-1M and Qwen2.5-14B-Instruct-1M, as well as the API-accessed model Qwen2.5-Turbo. Evaluations show that Qwen2.5-1M models have been greatly improved in long-context tasks without compromising performance in short-context scenarios. Specifically, the Qwen2.5-14B-Instruct-1M model significantly outperforms GPT-4o-mini in long-context tasks and supports contexts eight times longer.

Qwen2.5 Technical Report

In this report, we introduce Qwen2.5, a comprehensive series of large language models (LLMs) designed to meet diverse needs. Compared to previous iterations, Qwen 2.5 has been significantly improved during both the pre-training and post-training stages. In terms of pre-training, we have scaled the high-quality pre-training datasets from the previous 7 trillion tokens to 18 trillion tokens. This provides a strong foundation for common sense, expert knowledge, and reasoning capabilities. In terms of post-training, we implement intricate supervised finetuning with over 1 million samples, as well as multistage reinforcement learning. Post-training techniques enhance human preference, and notably improve long text generation, structural data analysis, and instruction following. To handle diverse and varied use cases effectively, we present Qwen2.5 LLM series in rich sizes. Open-weight offerings include base and instruction-tuned models, with quantized versions available. In addition, for hosted solutions, the proprietary models currently include two mixture-of-experts (MoE) variants: Qwen2.5-Turbo and Qwen2.5-Plus, both available from Alibaba Cloud Model Studio. Qwen2.5 has demonstrated top-tier performance on a wide range of benchmarks evaluating language understanding, reasoning, mathematics, coding, human preference alignment, etc. Specifically, the open-weight flagship Qwen2.5-72B-Instruct outperforms a number of open and proprietary models and demonstrates competitive performance to the state-of-the-art open-weight model, Llama-3-405B-Instruct, which is around 5 times larger. Qwen2.5-Turbo and Qwen2.5-Plus offer superior cost-effectiveness while performing competitively against GPT-4o-mini and GPT-4o respectively. Additionally, as the foundation, Qwen2.5 models have been instrumental in training specialized models such as Qwen2.5-Math, Qwen2.5-Coder, QwQ, and multimodal models.