arXiv subjects

Leilei Wang

Publications and source records attributed to Leilei Wang.

8 recordsLinked to original sources

arXiv2026

Controlling pair dynamics of rotating magnetic microparticles through radial and transverse interactions

Rotating magnetic microparticles are building blocks for field-driven assembly and microrobotic control. As the elementary interaction rule for larger assemblies, pair motion in these systems is governed not only by magnetic forcing, but also by hydrodynamic coupling and other long-range interactions. Here we develop a reduced framework for two synchronized rotating magnetic particles by resolving the interactions into radial components that change the interparticle distance and transverse components that rotate the line of centers. The competition between magnetic dipolar interaction, additional radial repulsion, and rotation-induced transverse coupling selects three pair-motion modes: rigid-body rotation, contact-separation rotation, and irreversible separation. We derive transition criteria for the rigid-body state, reversed orbital motion, and the separation boundary, and obtain an asymptotic solution for the separation dynamics. Lattice Boltzmann simulations of particles rotating near a wall provide a hydrodynamic realization of the model, in which inertial secondary flow generates radial repulsion and rotational flow produces transverse coupling. The resulting phase diagram in the physical $(\mathrm{Re},C_m)$ plane is consistent with the reduced-model predictions. These results provide design rules for programming elementary pair interactions in rotating magnetic-particle systems and may help guide the control of microrobotic assemblies.

cond-mat.soft↗

arXiv2026

SAGE: Multi-Agent Self-Evolution for LLM Reasoning

Reinforcement learning with verifiable rewards improves reasoning in large language models (LLMs), but many methods still rely on large human-labeled datasets. While self-play reduces this dependency, it often lacks explicit planning and strong quality control, limiting stability in long-horizon multi-step reasoning. We present SAGE (Self-evolving Agents for Generalized reasoning Evolution), a closed-loop framework where four agents: Challenger, Planner, Solver, and Critic, co-evolve from a shared LLM backbone using only a small seed set. The Challenger continuously generates increasingly difficult tasks; the Planner converts each task into a structured multi-step plan; and the Solver follows the plan to produce an answer, whose correctness is determined by external verifiers. The Critic scores and filters both generated questions and plans to prevent curriculum drift and maintain training signal quality, enabling stable self-training. Across mathematics and code-generation benchmarks, SAGE delivers consistent gains across model scales, improving the Qwen-2.5-7B model by 8.9% on LiveCodeBench and 10.7% on OlympiadBench.

cs.AI↗

arXiv2026

OV-DEIM: Real-time DETR-Style Open-Vocabulary Object Detection with GridSynthetic Augmentation

Real-time open-vocabulary object detection (OVOD) is essential for practical deployment in dynamic environments, where models must recognize a large and evolving set of categories under strict latency constraints. Current real-time OVOD methods are predominantly built upon YOLO-style models. In contrast, real-time DETR-based methods still lag behind in terms of inference latency, model lightweightness, and overall performance. In this work, we present OV-DEIM, an end-to-end DETR-style open-vocabulary detector built upon the recent DEIMv2 framework with integrated vision-language modeling for efficient open-vocabulary inference. We further introduce a simple query supplement strategy that improves Fixed AP without compromising inference speed. Beyond architectural improvements, we introduce GridSynthetic, a simple yet effective data augmentation strategy that composes multiple training samples into structured image grids. By exposing the model to richer object co-occurrence patterns and spatial layouts within a single forward pass, GridSynthetic mitigates the negative impact of noisy localization signals on the classification loss and improves semantic discrimination, particularly for rare categories. Extensive experiments demonstrate that OV-DEIM achieves state-of-the-art performance on open-vocabulary detection benchmarks, delivering superior efficiency and notable improvements on challenging rare categories. Code and pretrained models are available at https://github.com/wleilei/OV-DEIM.

cs.CV↗

arXiv2026

ForeDiffusion: Foresight-Conditioned Diffusion Policy via Future View Construction for Robot Manipulation

Diffusion strategies have advanced visual motor control by progressively denoising high-dimensional action sequences, providing a promising method for robot manipulation. However, as task complexity increases, the success rate of existing baseline models decreases considerably. Analysis indicates that current diffusion strategies are confronted with two limitations. First, these strategies only rely on short-term observations as conditions. Second, the training objective remains limited to a single denoising loss, which leads to error accumulation and causes grasping deviations. To address these limitations, this paper proposes Foresight-Conditioned Diffusion (ForeDiffusion), by injecting the predicted future view representation into the diffusion process. As a result, the policy is guided to be forward-looking, enabling it to correct trajectory deviations. Following this design, ForeDiffusion employs a dual loss mechanism, combining the traditional denoising loss and the consistency loss of future observations, to achieve the unified optimization. Extensive evaluation on the Adroit suite and the MetaWorld benchmark demonstrates that ForeDiffusion achieves an average success rate of 80% for the overall task, significantly outperforming the existing mainstream diffusion methods by 23% in complex tasks, while maintaining more stable performance across the entire tasks.

cs.RO↗

arXiv2025

Emergent Synchronization and Defect Dynamics in Confined Chiral Active Suspensions

Hydrodynamic interactions can generate rich emergent structures in active matter systems. Using large-scale hydrodynamic simulations, we demonstrate that hydrodynamic coupling alone can drive spontaneous self-organization across a hierarchy of spatial and temporal scales in confined suspensions of torque-driven particles at moderate Reynolds numbers. Spinners first self-assemble into dimers, which crystallize into a hexatic lattice and subsequently undergo a collective tilting instability. The resulting tilted dimers rotate and synchronize through hydrodynamic repulsion, which can be tuned by the Reynolds number. Upon synchronization, the polar director develops splay and bend deformations and nucleates topological defects with charges of $\pm1$. These defects induce long-wavelength concentration gradients and drive crystal vortex dynamics spanning hundreds of particle diameters. Our results reveal a purely hydrodynamic route to synchronization and defect-mediated dynamics in chiral active matter, without explicit alignment rules or interparticle forces.

cond-mat.soft↗

arXiv2024

RRWKV: Capturing Long-range Dependencies in RWKV

Owing to the impressive dot-product attention, the Transformers have been the dominant architectures in various natural language processing (NLP) tasks. Recently, the Receptance Weighted Key Value (RWKV) architecture follows a non-transformer architecture to eliminate the drawbacks of dot-product attention, where memory and computational complexity exhibits quadratic scaling with sequence length. Although RWKV has exploited a linearly tensor-product attention mechanism and achieved parallelized computations by deploying the time-sequential mode, it fails to capture long-range dependencies because of its limitation on looking back at previous information, compared with full information obtained by direct interactions in the standard transformer. Therefore, the paper devises the Retrospected Receptance Weighted Key Value (RRWKV) architecture via incorporating the retrospecting ability into the RWKV to effectively absorb information, which maintains memory and computational efficiency as well.

cs.CL↗

arXiv2024

Solution for CVPR 2024 UG2+ Challenge Track on All Weather Semantic Segmentation

In this report, we present our solution for the semantic segmentation in adverse weather, in UG2+ Challenge at CVPR 2024. To achieve robust and accurate segmentation results across various weather conditions, we initialize the InternImage-H backbone with pre-trained weights from the large-scale joint dataset and enhance it with the state-of-the-art Upernet segmentation method. Specifically, we utilize offline and online data augmentation approaches to extend the train set, which helps us to further improve the performance of the segmenter. As a result, our proposed solution demonstrates advanced performance on the test set and achieves 3rd position in this challenge.

cs.CV↗

arXiv2023

ESIE-BERT: Enriching Sub-words Information Explicitly with BERT for Joint Intent Classification and SlotFilling

Natural language understanding (NLU) has two core tasks: intent classification and slot filling. The success of pre-training language models resulted in a significant breakthrough in the two tasks. One of the promising solutions called BERT can jointly optimize the two tasks. We note that BERT-based models convert each complex token into multiple sub-tokens by wordpiece algorithm, which generates a mismatch between the lengths of the tokens and the labels. This leads to BERT-based models do not do well in label prediction which limits model performance improvement. Many existing models can be compatible with this issue but some hidden semantic information is discarded in the fine-tuning process. We address the problem by introducing a novel joint method on top of BERT which explicitly models the multiple sub-tokens features after wordpiece tokenization, thereby contributing to the two tasks. Our method can well extract the contextual features from complex tokens by the proposed sub-words attention adapter (SAA), which preserves overall utterance information. Additionally, we propose an intent attention adapter (IAA) to obtain the full sentence features to aid users to predict intent. Experimental results confirm that our proposed model is significantly improved on two public benchmark datasets. In particular, the slot filling F1 score is improved from 96.1 to 98.2 (2.1% absolute) on the Airline Travel Information Systems (ATIS) dataset.

cs.CL↗