arXiv subjects

Qianyun Zhang

Publications and source records attributed to Qianyun Zhang.

6 recordsLinked to original sources

arXiv2026

Self-evolving Agentic Customer Support System at LinkedIn

Enterprise support agents operate in rapidly changing environments where policies, product capabilities, and knowledge bases evolve continuously, making static assistants brittle and costly to maintain. We present LinkedIn's self-evolving agentic support system, which integrates retrieval-augmented generation with evolutionary auto-prompting and a modular, production-aligned evaluation framework to enable safe, continuous improvement without retraining foundation models. The system treats prompts, retrieval, and evaluation as a closed-loop, versioned workflow with operational guardrails. Offline simulations and ablations show clear quality gains over vanilla RAG and baseline agents, including reduced hallucinations and improved response completeness. In a two-week user-randomized A/B test on LinkedIn's production support traffic, the integrated self-evolved workflow increased QA self-serve by 9.0 percentage points, cancellation self-serve by 4.8 points, and routing accuracy by 30.6 points. These results demonstrate a practical path to scalable, self-evolving AI agents in real-world enterprise settings.

cs.AI↗

arXiv2024

Hardware-in-the-loop Simulation Testbed for Geomagnetic Navigation

Geomagnetic navigation leverages the ubiquitous Earth's magnetic signals to navigate missions, without dependence on GPS services or pre-stored geographic maps. It has drawn increasing attention and is promising particularly for long-range navigation into unexplored areas. Current geomagnetic navigation studies are still in the early stages with simulations and computational validations, without concrete efforts to develop cost-friendly test platforms that can empower deployment and experimental analysis of the developed approaches. This paper presents a hardware-in-the-loop simulation testbed to support geomagnetic navigation experimentation. Our testbed is dedicated to synthesizing geomagnetic field environment for the navigation. We develop the software in the testbed to simulate the dynamics of the navigation environment, and we build the hardware to generate the physical magnetic field, which follows and aligns with the simulated environment. The testbed aims to provide controllable magnetic field that can be used to experiment with geomagnetic navigation in labs, thus avoiding real and expensive navigation experiments, e.g., in the ocean, for validating navigation prototypes. We build the testbed with off-the-shelf hardware in an unshielded environment to reduce cost. We also develop the field generation control and hardware parameter optimization for quality magnetic field generation. We conduct a detailed performance analysis to show the quality of the field generation by the testbed, and we report the experimental results on performance indicators, including accuracy, uniformity, stability, and convergence of the generated field towards the target geomagnetic environment.

cs.RO↗

arXiv2024

A Secure and Efficient Distributed Semantic Communication System for Heterogeneous Internet of Things

Semantic communications are expected to improve the transmission efficiency in Internet of Things (IoT) networks. However, the distributed nature of networks and heterogeneity of devices challenge the secure utilization of semantic communication systems. In this paper, we develop a distributed semantic communication system that achieves the security and efficiency during update and usage phases. A blockchain-based trust scheme for update is designed to continuously train and synchronize the system in dynamic IoT environments. To improve the updating efficiency, we propose a flexible semantic coding method base on compressive semantic knowledge bases. It greatly reduces the amount of data shared among devices for system update, and realizes the flexible adjustment of the size of knowledge bases and the number of transmitted signal symbols in model training and inference stages. In the usage phase, a signature mechanism for lossy semantics is introduced to guarantee the integrity and authenticity of the transmitted semantics in lossy semantic communications. We further design a noise-aware differential privacy mechanism, which introduces optimized noise based on the different channel information available to heterogeneous devices. Experiments on text transmission tasks show that the proposed system achieves the protection of the integrity and privacy for exchanged semantics, and reduces the data to be transmitted in the update phase by about $35\%$ to $88\%$, and in the usage phase by $60\%$ compared with related works.

eess.SP↗

arXiv2022

Towards Tailored Models on Private AIoT Devices: Federated Direct Neural Architecture Search

Neural networks often encounter various stringent resource constraints while deploying on edge devices. To tackle these problems with less human efforts, automated machine learning becomes popular in finding various neural architectures that fit diverse Artificial Intelligence of Things (AIoT) scenarios. Recently, to prevent the leakage of private information while enable automated machine intelligence, there is an emerging trend to integrate federated learning and neural architecture search (NAS). Although promising as it may seem, the coupling of difficulties from both tenets makes the algorithm development quite challenging. In particular, how to efficiently search the optimal neural architecture directly from massive non-independent and identically distributed (non-IID) data among AIoT devices in a federated manner is a hard nut to crack. In this paper, to tackle this challenge, by leveraging the advances in ProxylessNAS, we propose a Federated Direct Neural Architecture Search (FDNAS) framework that allows for hardware-friendly NAS from non- IID data across devices. To further adapt to both various data distributions and different types of devices with heterogeneous embedded hardware platforms, inspired by meta-learning, a Cluster Federated Direct Neural Architecture Search (CFDNAS) framework is proposed to achieve device-aware NAS, in the sense that each device can learn a tailored deep learning model for its particular data distribution and hardware constraint. Extensive experiments on non-IID datasets have shown the state-of-the-art accuracy-efficiency trade-offs achieved by the proposed solution in the presence of both data and device heterogeneity.

cs.LG↗

arXiv2015

Scaling of liquid-drop impact craters in wet granular media

Combining high-speed photography with laser profilometry, we study the dynamics and the morphology of liquid-drop impact cratering in wet granular media---a ubiquitous phenomenon relevant to many important geological, agricultural, and industrial processes. By systematically investigating important variables such as impact energy, the size of impinging drops and the degree of liquid saturation in granular beds, we uncover a novel scaling for the size of impact craters. We show that this scaling can be explained by considering the balance between the inertia of impinging drops and the strength of impacted surface. Such a theoretical understanding confirms that the unique energy partition originally proposed for liquid-drop impact cratering in dry granular media also applies for impact cratering in wet granular media. Moreover, we demonstrate that compressive stresses, instead of shear stresses, control granular impact cratering. Our study enriches the picture of generic granular impact cratering and sheds light on the familiar phenomena of raindrop impacts in granular media.

cond-mat.soft↗

arXiv2014

Granular impact cratering by liquid drops: Understanding raindrop imprints through an analogy to asteroid strikes

When a granular material is impacted by a sphere, its surface deforms like a liquid yet it preserves a circular crater like a solid. Although the mechanism of granular impact cratering by solid spheres is well explored, our knowledge on granular impact cratering by liquid drops is still very limited. Here, by combining high-speed photography with high-precision laser profilometry, we investigate liquid-drop impact dynamics on granular surface and monitor the morphology of resulting impact craters. Surprisingly, we find that, despite the enormous energy and length difference, granular impact cratering by liquid drops follows the same energy scaling and reproduces the same crater morphology as that of asteroid impact craters. Inspired by this similarity, we integrate the physical insight from planetary sciences, the liquid marble model from fluid mechanics and the concept of jamming transition from granular physics into a simple theoretical framework that quantitatively describes all the main features of liquid-drop imprints in granular media. Our study sheds light on the mechanisms governing raindrop impacts on granular surfaces and reveals a remarkable analogy between familiar phenomena of raining and catastrophic asteroid strikes.

cond-mat.soft↗