arXiv subjects

Xinyu Pan

Publications and source records attributed to Xinyu Pan.

8 recordsLinked to original sources

arXiv2026

A Survey of Agent Memory in the Second Half: Towards Self-Evolving and Long-Horizon Agents

Research in artificial intelligence is shifting from model innovations and benchmark scores towards problem definition and rigorous real-world evaluation. As the field enters the "second half," the central challenge becomes real utility in long-horizon, dynamic, and user-dependent settings such as agentic coding, deep research, and computer use, where LLM-based agents face context explosion beyond fixed context windows and must continuously accumulate, manage, and selectively reuse information across extended interactions. Memory, with hundreds of papers released in 2025, therefore emerges as the critical solution to fill this utility gap. Beyond passive storage, memory is increasingly the substrate through which agents self-evolve: short-term memory gates which experiences are perceived and abstracted during execution, while long-term memory consolidates them into reusable knowledge and skills, forming the loop through which agents improve from their own experience. In this survey, we provide a unified view of foundation agent memory along three dimensions: memory substrate (internal parametric state and external retrieval-augmented stores), cognitive mechanism (sensory, working, episodic, semantic, and procedural), and memory subject (user-centric personalization and agent-centric experience). We then analyze how memory is operated under single- and multi-agent topologies and highlight learning policies over memory operations, showing how memory management itself is becoming a trainable capability spanning reinforcement-learned context curation, experience consolidation at decision time, and the emerging ecosystem of portable, shareable agent skills. Finally, we review evaluation benchmarks and metrics for memory utility, and outline open challenges and future directions.

cs.CL↗

arXiv2025

LLM4AMC: Adapting Large Language Models for Adaptive Modulation and Coding

Adaptive modulation and coding (AMC) is a key technology in 5G new radio (NR), enabling dynamic link adaptation by balancing transmission efficiency and reliability based on channel conditions. However, traditional methods often suffer from performance degradation due to the aging issues of channel quality indicator (CQI). Recently, the emerging capabilities of large language models (LLMs) in contextual understanding and temporal modeling naturally align with the dynamic channel adaptation requirements of AMC technology. Leveraging pretrained LLMs, we propose a channel quality prediction method empowered by LLMs to optimize AMC, termed LLM4AMC. We freeze most parameters of the LLM and fine-tune it to fully utilize the knowledge acquired during pretraining while better adapting it to the AMC task. We design a network architecture composed of four modules, a preprocessing layer, an embedding layer, a backbone network, and an output layer, effectively capturing the time-varying characteristics of channel quality to achieve accurate predictions of future channel conditions. Simulation experiments demonstrate that our proposed method significantly improves link performance and exhibits potential for practical deployment.

eess.SP↗

arXiv2025

PSDiffusion: Harmonized Multi-Layer Image Generation via Layout and Appearance Alignment

Transparent image layer generation plays a significant role in digital art and design workflows. Existing methods typically decompose transparent layers from a single RGB image using a set of tools or generate multiple transparent layers sequentially. Despite some promising results, these methods often limit their ability to model global layout, physically plausible interactions, and visual effects such as shadows and reflections with high alpha quality due to limited shared global context among layers. To address this issue, we propose PSDiffusion, a unified diffusion framework that leverages image composition priors from pre-trained image diffusion model for simultaneous multi-layer text-to-image generation. Specifically, our method introduces a global layer interaction mechanism to generate layered images collaboratively, ensuring both individual layer quality and coherent spatial and visual relationships across layers. We include extensive experiments on benchmark datasets to demonstrate that PSDiffusion is able to outperform existing methods in generating multi-layer images with plausible structure and enhanced visual fidelity.

cs.CV↗

arXiv2022

The Hydrogen Intensity and Real-time Analysis eXperiment: 256-Element Array Status and Overview

The Hydrogen Intensity and Real-time Analysis eXperiment (HIRAX) is a radio interferometer array currently in development, with an initial 256-element array to be deployed at the South African Radio Astronomy Observatory (SARAO) Square Kilometer Array (SKA) site in South Africa. Each of the 6m, $f/0.23$ dishes will be instrumented with dual-polarisation feeds operating over a frequency range of 400-800 MHz. Through intensity mapping of the 21 cm emission line of neutral hydrogen, HIRAX will provide a cosmological survey of the distribution of large-scale structure over the redshift range of $0.775 < z < 2.55$ over $\sim$15,000 square degrees of the southern sky. The statistical power of such a survey is sufficient to produce $\sim$7 percent constraints on the dark energy equation of state parameter when combined with measurements from the Planck satellite. Additionally, HIRAX will provide a highly competitive platform for radio transient and HI absorber science while enabling a multitude of cross-correlation studies. In this paper, we describe the science goals of the experiment, overview of the design and status of the sub-components of the telescope system, and describe the expected performance of the initial 256-element array as well as the planned future expansion to the final, 1024-element array.

astro-ph.IM↗

arXiv2020

Evolutionary Stochastic Policy Distillation

Solving the Goal-Conditioned Reward Sparse (GCRS) task is a challenging reinforcement learning problem due to the sparsity of reward signals. In this work, we propose a new formulation of GCRS tasks from the perspective of the drifted random walk on the state space, and design a novel method called Evolutionary Stochastic Policy Distillation (ESPD) to solve them based on the insight of reducing the First Hitting Time of the stochastic process. As a self-imitate approach, ESPD enables a target policy to learn from a series of its stochastic variants through the technique of policy distillation (PD). The learning mechanism of ESPD can be considered as an Evolution Strategy (ES) that applies perturbations upon policy directly on the action space, with a SELECT function to check the superiority of stochastic variants and then use PD to update the policy. The experiments based on the MuJoCo robotics control suite show the high learning efficiency of the proposed method.

cs.LG↗

arXiv2019

Observing information backflow from controllable non-Markovian multi-channels in diamond

Any realistic quantum system is inevitably subject to an external environment. This environment makes the open-system dynamics significant for many quantum tech-nologies, such as entangled-state engineering, quantum simulation, and quantum sensing. The information flow of a system to its environment usually induces a Markovian process, while the backflow of information from the environment exhibits non-Markovianity. The practical environment, usually consisting of a large number of degrees of freedom, is hard to control, despite some attempts on controllable transitions from Markovian to non-Markovian dynamics. Here, we experimentally demonstrate the engineering of multiple dissipative channels by controlling the adjacent nuclear spins of a nitrogen-vacancy centre in diamond. With a controllable non-Markovian dynamics of the open system, we observe that the quantum Fisher information flows to and from the environment using different noisy channels. Our work contributes to the developments of both noisy quantum metrology and quantum open systems from the view points of metrologically useful entanglement.

quant-ph↗

arXiv2018

Quantization Mimic: Towards Very Tiny CNN for Object Detection

In this paper, we propose a simple and general framework for training very tiny CNNs for object detection. Due to limited representation ability, it is challenging to train very tiny networks for complicated tasks like detection. To the best of our knowledge, our method, called Quantization Mimic, is the first one focusing on very tiny networks. We utilize two types of acceleration methods: mimic and quantization. Mimic improves the performance of a student network by transfering knowledge from a teacher network. Quantization converts a full-precision network to a quantized one without large degradation of performance. If the teacher network is quantized, the search scope of the student network will be smaller. Using this feature of the quantization, we propose Quantization Mimic. It first quantizes the large network, then mimic a quantized small network. The quantization operation can help student network to better match the feature maps from teacher network. To evaluate our approach, we carry out experiments on various popular CNNs including VGG and Resnet, as well as different detection frameworks including Faster R-CNN and R-FCN. Experiments on Pascal VOC and WIDER FACE verify that our Quantization Mimic algorithm can be applied on various settings and outperforms state-of-the-art model acceleration methods given limited computing resouces.

cs.CV↗

arXiv2010

Laser and Microwave Excitations of Rabi Oscillations of a Single Nitrogen-Vacancy Electron Spin in Diamond

A collapse and revival shape of Rabi oscillations of a single Nitrogen-Vacancy (NV) center electron spin has been observed in diamond at room temperature. Because of hyperfine interaction between the host 14N nuclear spin and NV center electron spin, different orientation of the 14N nuclear spin leads to a triplet splitting of the transition between the ground ms=0 and excited states ms=1. Microwave can excite the three transitions equally to induce three independent nutations and the shape of Rabi oscillations is a combination of the three nutations. This result provides an innovative view of electron spin oscillations in diamond.

quant-ph↗