arXiv subjects

Yujie Zheng

Publications and source records attributed to Yujie Zheng.

9 recordsLinked to original sources

arXiv2026

He3-Seeker: Robotic Information Planning for Lunar Helium-3 Distribution Mapping

Lunar helium-3 is a highly valuable strategic resource, pivotal to the advancement of both deep-space exploration and space mining. Existing lunar helium-3 exploration methodologies rely primarily on indirect measurements via remote sensing, which are often characterized by limited precision, low reliability, and insufficient spatial resolution. In this paper, we introduce He3-Seeker, an active robotic exploration method for helium-3 distribution mapping. First, we provide a formal definition of the active helium-3 exploration problem. Subsequently, we developed the He3-Seeker framework, which is conceptually based on multi-point drilling, sampling, and in situ analysis. In particular, we use robotic information planning (RIP) to guide autonomous robot navigation and active sensing. Additionally, to thoroughly evaluate the proposed algorithm, we introduce a reliable method for generating reference data of lunar helium-3 distribution based on low-resolution orbital remote sensing measurements. Simulation experiments verify that He3-Seeker achieves both rapid and high-fidelity mapping of helium-3 distribution, providing a reliable solution for resource exploration tasks. Our code and simulation environment will be publicly accessible at https://github.com/OpenSpace-Lab/He3-Seeker.

cs.RO↗

arXiv2026

Beyond Retrieval: Learning Compact User Representations for Scalable LLM Personalization

Personalizing large language models requires adapting model behavior to individual users while preserving robustness and deployment-scale efficiency. Existing approaches typically personalize LLMs either at the input level, by retrieving user histories or constructing profile prompts, or at the parameter level, by maintaining user-specific parameter-efficient modules. The former makes personalization sensitive to retrieval quality and prompt design, whereas the latter incurs storage and maintenance costs that grow with the user population. To address these limitations, we propose TAP-PER (Temporal Attentive Prefix for PERsonalization), a prefix-based framework that encodes user preferences as learnable representations, avoiding the serialization of user histories into prompts and replacing heavy per-user adapters with lightweight user-state prefix embeddings. Inspired by personalized recommendation systems, TAP-PER decomposes user modeling into user-state and query-conditioned components, and incorporates temporal signals to capture the evolving nature of user interests. Experiments on six LaMP tasks show that TAP-PER consistently outperforms prompt-based and model-based baselines across classification, rating, and generation settings. Moreover, TAP-PER uses 130x fewer per-user parameters than OPPU and roughly half the total parameter footprint of PER-PCS at the 1,000-user scale, demonstrating scalable personalization without prompt-serialized histories or heavy per-user adapters.

cs.IR↗

arXiv2026

A Task-State Representation for Long-Horizon Mobile GUI Agents

While long-horizon mobile GUI agents typically rely on thought-action-observation loops, they struggle to separate persistent task states from transient screen observations. As execution histories grow, this entanglement imposes a severe context burden, causing agents to forget initial requirements, hallucinate progress, or repeatedly interact with stale interfaces. To address this, we introduce Task-State Representation (TSR), a training-free framework that explicitly decouples task state from sensory input. Acting as a lightweight external wrapper, TSR maintains three structured components: a global instruction summary, a dynamic progress tracker for subgoals, and a transition-aware action verifier. By continuously updating through pre- and post-action visual comparisons, TSR effectively guides the agent's reasoning without requiring architectural modifications. Experiments across four mobile GUI benchmarks validate TSR's effectiveness, yielding up to a 12 absolute point increase in success rate on complex cross-application and memory-intensive tasks.

cs.CL↗

arXiv2026

Towards Cold-Start Drafting and Continual Refining: A Value-Driven Memory Approach with Application to NPU Kernel Synthesis

Deploying Large Language Models to data-scarce programming domains poses significant challenges, particularly for kernel synthesis on emerging Domain-Specific Architectures where a "Data Wall" limits available training data. While models excel on data-rich platforms like CUDA, they suffer catastrophic performance drops on data-scarce ecosystems such as NPU programming. To overcome this cold-start barrier without expensive fine-tuning, we introduce EvoKernel, a self-evolving agentic framework that automates the lifecycle of kernel synthesis from initial drafting to continual refining. EvoKernel addresses this by formulating the synthesis process as a memory-based reinforcement learning task. Through a novel value-driven retrieval mechanism, it learns stage-specific Q-values that prioritize experiences based on their contribution to the current objective, whether bootstrapping a feasible draft or iteratively refining latency. Furthermore, by enabling cross-task memory sharing, the agent generalizes insights from simple to complex operators. By building an NPU variant of KernelBench and evaluating on it, EvoKernel improves frontier models' correctness from 11.0% to 83.0% and achieves a median speedup of 3.60x over initial drafts through iterative refinement. This demonstrates that value-guided experience accumulation allows general-purpose models to master the kernel synthesis task on niche hardware ecosystems. Our official page is available at https://evokernel.zhuo.li.

cs.LG↗

arXiv2026

MemRL: Self-Evolving Agents via Runtime Reinforcement Learning on Episodic Memory

The hallmark of human intelligence is the self-evolving ability to master new skills by learning from past experiences. However, current AI agents struggle to emulate this self-evolution: fine-tuning is computationally expensive and prone to catastrophic forgetting, while existing memory-based methods rely on passive semantic matching that often retrieves noise. To address these challenges, we propose MemRL, a non-parametric approach that evolves via reinforcement learning on episodic memory. By decoupling stable reasoning from plastic memory, MemRL employs a Two-Phase Retrieval mechanism to filter noise and identify high-utility strategies through environmental feedback. Extensive experiments on HLE, BigCodeBench, ALFWorld, and Lifelong Agent Bench demonstrate that MemRL significantly outperforms state-of-the-art baselines, confirming that MemRL effectively reconciles the stability-plasticity dilemma, enabling continuous runtime improvement without weight updates. Code is available at https://github.com/MemTensor/MemRL.

cs.CL↗

arXiv2025

ML-Master: Towards AI-for-AI via Integration of Exploration and Reasoning

As AI capabilities advance toward and potentially beyond human-level performance, a natural transition emerges where AI-driven development becomes more efficient than human-centric approaches. A promising pathway toward this transition lies in AI-for-AI (AI4AI), which leverages AI techniques to automate and optimize the design, training, and deployment of AI systems themselves. While LLM-based agents have shown the potential to realize AI4AI, they are often unable to fully leverage the experience accumulated by agents during the exploration of solutions in the reasoning process, leading to inefficiencies and suboptimal performance. To address this limitation, we propose ML-Master, a novel AI4AI agent that seamlessly integrates exploration and reasoning by employing a selectively scoped memory mechanism. This approach allows ML-Master to efficiently combine diverse insights from parallel solution trajectories with analytical reasoning, guiding further exploration without overwhelming the agent with excessive context. We evaluate ML-Master on the MLE-Bench, where it achieves a 29.3% average medal rate, significantly surpassing existing methods, particularly in medium-complexity tasks, while accomplishing this superior performance within a strict 12-hour time constraint-half the 24-hour limit used by previous baselines. These results demonstrate ML-Master's potential as a powerful tool for advancing AI4AI.

cs.AI↗

arXiv2025

Polarization-Analyzed Small-Angle Neutron Scattering with an $\textit{in-situ}$ $^{3}$He neutron spin filter at the China Spallation Neutron Source

Polarization-analyzed small-angle neutron scattering (PASANS) is an advanced technique that enables the selective investigation of magnetic scattering phenomena in magnetic materials and distinguishes coherent scattering obscured by incoherent backgrounds, making it particularly valuable for cutting-edge research. The successful implementation of PASANS in China was achieved for the first time at the newly commissioned Very Small Angle Neutron Scattering (VSANS) instrument at the China Spallation Neutron Source (CSNS). This technique employs a combination of a double-V cavity supermirror polarizer and a radio frequency (RF) neutron spin flipper to manipulate the polarization of the incident neutrons. The scattered neutron polarization is stably analyzed by a specially designed $\textit{in-situ}$ optical pumping $^{3}$He neutron spin filter, which covers a spatially symmetric scattering angle coverage of about 4.8 $^{\circ}$. A comprehensive PASANS data reduction method, aimed at pulsed neutron beams, has been established and validated with a silver behenate powder sample, indicating a maximum momentum transfer coverage of approximately 0.25 Å $^{-1}$.

physics.ins-det↗

arXiv2020

DC electricity generation from dynamic polarized water-semiconductor interface

Liquid electricity generator and hydrovoltaic technology have received numerous attentions, which can be divided into horizontal movement generator and vertical movement generator. The horizontal movement generator is limited for powering the integrated and miniaturized energy chip as the current output direction is depending on the moving direction of the water droplet, which means a sustainable and continuous direct-current (DC) electricity output can be hardly achieved because of the film of limited length. On the other hand, the existing vertical movement generators include triboelectricity or humidity gradient-based liquid electricity generator, where the liquid or water resource must be sustainably supplied to ensure continuous current output. Herein, we have designed an integratable vertical generator by sandwiching water droplets with semiconductor and metal, such as graphene or aluminum. This generator, named as polarized liquid molecular generator (PLMG), directly converts the lateral kinetic energy of water droplet into vertical DC electricity with an output voltage of up to ~1.0 V from the dynamic water-semiconductor interface. The fundamental discovery of PLMG is related to the non-symmetric structure of liquid molecules, such as water and alcohols, which can be polarized under the guidance of built-in field caused by the Fermi level difference between metal and semiconductor, while the symmetric liquid molecules cannot produce any electricity on the opposite. Integratable PLMG with a large output power of ~90 nW and voltage of ~2.7 V has been demonstrated, meanwhile its small internal resistance of ~250 kilohm takes a huge advantage in resistance matching with the impedance of electron components. The PLMG shows potential application value in the Internet of Things (IoTs) after proper miniaturization and integration.

physics.app-ph↗

arXiv2016

Automated Fourier space region-recognition filtering for off-axis digital holographic microscopy

Automated label-free quantitative imaging of biological samples can greatly benefit high throughput diseases diagnosis. Digital holographic microscopy (DHM) is a powerful quantitative label-free imaging tool that retrieves structural details of cellular samples non-invasively. In off-axis DHM, a proper spatial filtering window in Fourier space is crucial to the quality of reconstructed phase image. Here we describe a region-recognition approach that combines shape recognition with an iterative thresholding to extracts the optimal shape of frequency components. The region recognition technique offers fully automated adaptive filtering that can operate with a variety of samples and imaging conditions. When imaging through optically scattering biological hydrogel matrix, the technique surpasses previous histogram thresholding techniques without requiring any manual intervention. Finally, we automate the extraction of the statistical difference of optical height between malaria parasite infected and uninfected red blood cells. The method described here pave way to greater autonomy in automated DHM imaging for imaging live cell in thick cell cultures.

physics.optics↗