arXiv subjects

Jiawei Yuan

Publications and source records attributed to Jiawei Yuan.

10 recordsLinked to original sources

arXiv2026

Kwai Keye-VL-2.0 Technical Report

We introduce Kwai Keye-VL-2.0-30B-A3B, an open-source Mixture-of-Experts (MoE) multimodal foundation model designed to advance long-video understanding and agentic intelligence. To address the challenges of ultra-long contexts, information redundancy, and prohibitive computational costs inherent in hour-level videos, Keye-VL-2.0 is the first to adapt DeepSeek Sparse Attention (DSA) to GQA-based multimodal architectures, enabling lossless 256K context processing while capturing critical frames and long-range temporal dependencies. This architecture is underpinned by a highly optimized training and inference infrastructure, including scalable video I/O, heterogeneous ViT-LM parallelism, and custom DSA kernels that significantly maximize throughput and minimize computational overhead. Furthermore, to overcome the algorithmic dilemma of catastrophic forgetting during multi-task alignment, we introduce Cross-Modal Multi-Teacher On-Policy Distillation (MOPD) paired with Context-RL and Video-RL. By distilling dense token-level teacher feedback from on-policy rollouts back into the MoE backbone, which activates only 3B parameters, Keye-VL-2.0 natively empowers advanced agent collaboration across Code, Tool, and Search scenarios with multimodal self-correction. Extensive evaluations across video understanding, temporal grounding, reasoning, STEM, and agent benchmarks demonstrate that Keye-VL-2.0-30B-A3B achieves state-of-the-art performance among models of similar scale, particularly excelling in fine-grained temporal localization on TimeLens and long-video comprehension on Video-MME-v2 and LongVideoBench. We release our model checkpoints to accelerate community progress toward scalable and robust multimodal agentic applications.

cs.CV↗

arXiv2026

WirelessSenseLLM: Zero-Shot Human Activity Understanding by Bridging Wireless Signals and Human Language

There is growing interest in enabling wireless sensing systems to interpret human motion from unsegmented wireless signals; however, existing CSI-based applications rely heavily on accurate signal segmentation and predefined action labels, limiting their applicability in zero-shot scenarios. We present WirelessSenseLLM, a language-driven framework that leverages large language models (LLMs) to enable zero-shot human motion understanding from unsegmented Wi-Fi Channel State Information (CSI). To bridge the modality gap between time-series CSI and discrete language representations, we introduce a CSI-to-Language Adapter and a cross-modal projection mechanism that maps CSI features into a language-aligned semantic space. This design enables the generation of fine-grained natural language descriptions of sequential and overlapping human motions, supporting downstream reasoning without segmented training data. We address two core technical challenges: modality mismatch between CSI features and language embeddings, and overlapping actions in unsegmented CSI streams. Extensive experiments demonstrate strong performance in zero-shot action understanding (92% accuracy and 91% F1-score), language-based reasoning quality (30% factual and 15% reasoning improvements), and multi-person motion explanation with an average 12.33% improvement over prior methods. These results highlight WirelessSenseLLM's effectiveness for robust and interpretable human motion understanding from CSI signals.

cs.NI↗

arXiv2026

Ro-SLM: Onboard Small Language Models for Robot Task Planning and Operation Code Generation

Recent advances in large language models (LLMs) provide robots with contextual reasoning abilities to comprehend human instructions. Yet, current LLM-enabled robots typically depend on cloud-based models or high-performance computing infrastructure, which limit their deployment on robots under unreliable internet environments or with constrained computational resources, such as UAVs and small ground vehicles. Thus, deploying fine-tuned small language models (SLMs) that support onboard deployment offers a promising alternative. This paper introduces Ro-SLM, a framework that enables reliable SLM-driven robot operation by distilling LLMs' knowledge and reasoning. Ro-SLM starts from dataset synthesis by leveraging LLMs to generate diverse task instructions, produce corresponding ground truth code with minimal human assistance, and augment instructions into real-world application scenarios. Ro-SLM is then fine-tuned with the dataset, in which LLM serves as a reward function to guide the training. Extensive experiments on UAV operation tasks demonstrate that Ro-SLM improves the performance of SLM from being incapable of supporting robotic task planning and code generation to achieving performance that approaches LLM.

cs.RO↗

arXiv2026

LLM-Driven Corrective Robot Operation Code Generation with Static Text-Based Simulation

Recent advances in Large language models (LLMs) have demonstrated their promising capabilities of generating robot operation code to enable LLM-driven robots. To enhance the reliability of operation code generated by LLMs, corrective designs with feedback from the observation of executing code have been increasingly adopted in existing research. However, the code execution in these designs relies on either a physical experiment or a customized simulation environment, which limits their deployment due to the high configuration effort of the environment and the potential long execution time. In this paper, we explore the possibility of directly leveraging LLM to enable static simulation of robot operation code, and then leverage it to design a new reliable LLM-driven corrective robot operation code generation framework. Our framework configures the LLM as a static simulator with enhanced capabilities that reliably simulate robot code execution by interpreting actions, reasoning over state transitions, analyzing execution outcomes, and generating semantic observations that accurately capture trajectory dynamics. To validate the performance of our framework, we performed experiments on various operation tasks for different robots, including UAVs and small ground vehicles. The experiment results not only demonstrated the high accuracy of our static text-based simulation but also the reliable code generation of our LLM-driven corrective framework, which achieves a comparable performance with state-of-the-art research while does not rely on dynamic code execution using physical experiments or simulators.

cs.RO↗

arXiv2026

Physics Encoded Spatial and Temporal Generative Adversarial Network for Tropical Cyclone Image Super-resolution

High-resolution satellite imagery is indispensable for tracking the genesis, intensification, and trajectory of tropical cyclones (TCs). However, existing deep learning-based super-resolution (SR) methods often treat satellite image sequences as generic videos, neglecting the underlying atmospheric physical laws governing cloud motion. To address this, we propose a Physics Encoded Spatial and Temporal Generative Adversarial Network (PESTGAN) for TC image super-resolution. Specifically, we design a disentangled generator architecture incorporating a PhyCell module, which approximates the vorticity equation via constrained convolutions and encodes the resulting approximate physical dynamics as implicit latent representations to separate physical dynamics from visual textures. Furthermore, a dual-discriminator framework is introduced, employing a temporal discriminator to enforce motion consistency alongside spatial realism. Experiments on the Digital Typhoon dataset for 4$\times$ upscaling demonstrate that PESTGAN establishes a better performance in structural fidelity and perceptual quality. While maintaining competitive pixel-wise accuracy compared to existing approaches, our method significantly excels in reconstructing meteorologically plausible cloud structures with superior physical fidelity.

cs.CV↗

arXiv2026

NeuroGenPoisoning: Neuron-Guided Attacks on Retrieval-Augmented Generation of LLM via Genetic Optimization of External Knowledge

Retrieval-Augmented Generation (RAG) empowers Large Language Models (LLMs) to dynamically integrate external knowledge during inference, improving their factual accuracy and adaptability. However, adversaries can inject poisoned external knowledge to override the model's internal memory. While existing attacks iteratively manipulate retrieval content or prompt structure of RAG, they largely ignore the model's internal representation dynamics and neuron-level sensitivities. The underlying mechanism of RAG poisoning has not been fully studied and the effect of knowledge conflict with strong parametric knowledge in RAG is not considered. In this work, we propose NeuroGenPoisoning, a novel attack framework that generates adversarial external knowledge in RAG guided by LLM internal neuron attribution and genetic optimization. Our method first identifies a set of Poison-Responsive Neurons whose activation strongly correlates with contextual poisoning knowledge. We then employ a genetic algorithm to evolve adversarial passages that maximally activate these neurons. Crucially, our framework enables massive-scale generation of effective poisoned RAG knowledge by identifying and reusing promising but initially unsuccessful external knowledge variants via observed attribution signals. At the same time, Poison-Responsive Neurons guided poisoning can effectively resolves knowledge conflict. Experimental results across models and datasets demonstrate consistently achieving high Population Overwrite Success Rate (POSR) of over 90% while preserving fluency. Empirical evidence shows that our method effectively resolves knowledge conflict.

cs.AI↗

arXiv2025

Large Language Model-Driven Closed-Loop UAV Operation with Semantic Observations

Recent advances in large Language Models (LLMs) have revolutionized mobile robots, including unmanned aerial vehicles (UAVs), enabling their intelligent operation within Internet of Things (IoT) ecosystems. However, LLMs still face challenges from logical reasoning and complex decision-making, leading to concerns about the reliability of LLM-driven UAV operations in IoT applications. In this paper, we propose a closed-loop LLM-driven UAV operation code generation framework that enables reliable UAV operations powered by effective feedback and refinement using two LLM modules, i.e., a Code Generator and an Evaluator. Our framework transforms numerical state observations from UAV operations into semantic trajectory descriptions to enhance the evaluator LLM's understanding of UAV dynamics for precise feedback generation. Our framework also enables a simulation-based refinement process, and hence eliminates the risks to physical UAVs caused by incorrect code execution during the refinement. Extensive experiments on UAV control tasks with different complexities are conducted. The experimental results show that our framework can achieve reliable UAV operations using LLMs, which significantly outperforms baseline methods in terms of success rate and completeness with the increase of task complexity.

cs.RO↗

arXiv2025

GSCE: A Prompt Framework with Enhanced Reasoning for Reliable LLM-driven Drone Control

The integration of Large Language Models (LLMs) into robotic control, including drones, has the potential to revolutionize autonomous systems. Research studies have demonstrated that LLMs can be leveraged to support robotic operations. However, when facing tasks with complex reasoning, concerns and challenges are raised about the reliability of solutions produced by LLMs. In this paper, we propose a prompt framework with enhanced reasoning to enable reliable LLM-driven control for drones. Our framework consists of novel technical components designed using Guidelines, Skill APIs, Constraints, and Examples, namely GSCE. GSCE is featured by its reliable and constraint-compliant code generation. We performed thorough experiments using GSCE for the control of drones with a wide level of task complexities. Our experiment results demonstrate that GSCE can significantly improve task success rates and completeness compared to baseline approaches, highlighting its potential for reliable LLM-driven autonomous drone systems.

cs.RO↗

arXiv2019

CPAR: Cloud-Assisted Privacy-preserving Image Annotation with Randomized KD-Forest

With the explosive growth in the number of pictures taken by smartphones, organizing and searching pictures has become important tasks. To efficiently fulfill these tasks, the key enabler is annotating images with proper keywords, with which keyword-based searching and organizing become available for images. Currently, smartphones usually synchronize photo albums with cloud storage platforms, and have their images annotated with the help of cloud computing. However, the "offloading-to-cloud" solution may cause privacy breach, since photos from smart photos contain various sensitive information. For privacy protection, existing research made effort to support cloud-based image annotation on encrypted images by utilizing cryptographic primitives. Nevertheless, for each annotation, it requires the cloud to perform linear checking on the large-scale encrypted dataset with high computational cost. This paper proposes a cloud-assisted privacy-preserving image annotation with randomized kd-forest, namely CPAR. With CPAR, users are able to automatically assign keywords to their images by leveraging the power of cloud with privacy protected. CPAR proposes a novel privacy-preserving randomized kd-forest structure, which significantly improves the annotation performance compared with existing research. Thorough analysis is carried out to demonstrate the security of CPAR. Experimental evaluation on the well-known IAPR TC-12 dataset validates the efficiency and effectiveness of CPAR.

cs.CR↗

arXiv2019

LEP-CNN: A Lightweight Edge Device Assisted Privacy-preserving CNN Inference Solution for IoT

Supporting convolutional neural network (CNN) inference on resource-constrained IoT devices in a timely manner has been an outstanding challenge for emerging smart systems. To mitigate the burden on IoT devices, the prevailing solution is to offload the CNN inference task, which is usually composed of billions of operations, to public cloud. However, the "offloading-to-cloud" solution may cause privacy breach while moving sensitive data to cloud. For privacy protection, the research community has resorted to advanced cryptographic primitives and approximation techniques to support CNN inference on encrypted data. Consequently, these attempts cause impractical computational overhead on IoT devices and degrade the performance of CNNs. Moreover, relying on the remote cloud can cause additional network latency and even make the system dysfunction when network connection is off. We proposes an extremely lightweight edge device assisted private CNN inference solution for IoT devices, namely LEP-CNN. The main design of LEP-CNN is based on a novel online/offline encryption scheme. The decryption of LEP-CNN is pre-computed offline via utilizing the linear property of the most time-consuming operations of CNNs. As a result, LEP-CNN allows IoT devices to securely offload over 99% CNN operations, and edge devices to execute CNN inference on encrypted data as efficient as on plaintext. LEP-CNN also provides an integrity check option to help IoT devices detect error results with a successful rate over 99%. Experiments on AlexNet show that LEP-CNN can speed up the CNN inference for more than 35 times for resource constrained IoT devices. A homomorphic encryption based AlexNet using CryptoNets is implemented to compare with LEP-CNN to demonstrate that LEP-CNN has a better performance than homomorphic encryption based privacy preserving neural networks under time-sensitive scenarios.

cs.CR↗