arXiv subjects

Haoming Sun

Publications and source records attributed to Haoming Sun.

3 recordsLinked to original sources

arXiv2026

How Implicit Bias Accumulates and Propagates in LLM Long-term Memory

Long-term memory mechanisms enable Large Language Models (LLMs) to maintain continuity and personalization across extended interaction lifecycles, but they also introduce new and underexplored risks related to fairness. In this work, we study how implicit bias, defined as subtle statistical prejudice, accumulates and propagates within LLMs equipped with long-term memory. To support systematic analysis, we introduce the Decision-based Implicit Bias (DIB) Benchmark, a large-scale dataset comprising 3,776 decision-making scenarios across nine social domains, designed to quantify implicit bias in long-term decision processes. Using a realistic long-horizon simulation framework, we evaluate six state-of-the-art LLMs integrated with three representative memory architectures on DIB and demonstrate that LLMs' implicit bias does not remain static but intensifies over time and propagates across unrelated domains. We further analyze mitigation strategies and show that a static system-level prompting baseline provides limited and short-lived debiasing effects. To address this limitation, we propose Dynamic Memory Tagging (DMT), an agentic intervention that enforces fairness constraints at memory write time. Extensive experimental results show that DMT substantially reduces bias accumulation and effectively curtails cross-domain bias propagation.

cs.LG↗

arXiv2024

Self-assessment, Exhibition, and Recognition: a Review of Personality in Large Language Models

As large language models (LLMs) appear to behave increasingly human-like in text-based interactions, more and more researchers become interested in investigating personality in LLMs. However, the diversity of psychological personality research and the rapid development of LLMs have led to a broad yet fragmented landscape of studies in this interdisciplinary field. Extensive studies across different research focuses, different personality psychometrics, and different LLMs make it challenging to have a holistic overview and further pose difficulties in applying findings to real-world applications. In this paper, we present a comprehensive review by categorizing current studies into three research problems: self-assessment, exhibition, and recognition, based on the intrinsic characteristics and external manifestations of personality in LLMs. For each problem, we provide a thorough analysis and conduct in-depth comparisons of their corresponding solutions. Besides, we summarize research findings and open challenges from current studies and further discuss their underlying causes. We also collect extensive publicly available resources to facilitate interested researchers and developers. Lastly, we discuss the potential future research directions and application scenarios. Our paper is the first comprehensive survey of up-to-date literature on personality in LLMs. By presenting a clear taxonomy, in-depth analysis, promising future directions, and extensive resource collections, we aim to provide a better understanding and facilitate further advancements in this emerging field.

cs.CL↗

arXiv2023

Rotating spokes, potential hump and modulated ionization in radio frequency magnetron discharges

In this work, the rotating spoke mode in the radio frequency (RF) magnetron discharge, which features the potential hump and the RF-modulated ionization, is observed and analyzed by means of the two dimensional axial-azimuthal (z-y) particle-in-cell/Monte Carlo collision method. The kinetic model combined with the linear analysis of the perturbation reveals that the cathode sheath (axial) electric field $E_z$ triggers the gradient drift instability (GDI), deforming the local potential until the instability condition is not fulfilled and the fluctuation growth stops in which moment the instability becomes saturated. The potential deformation consequently leads to the formation of the potential hump, surrounding which the azimuthal electric field $E_y$ is present. The saturation level of $E_y$ is found to be synchronized with and proportional to the time-changing voltage applied at the cathode, resulting in the RF-modulation of the electron heating in the $E_y$ due to $\nabla B$ drift. In the instability saturated stage, it is shown that the rotation velocity and direction of the spoke present in the simulations agree well with the experimental observation. In the instability linear stage, the instability mode wavelength and the growth rate are also found to be in good agreement with the prediction of the GDI linear fluid theory.

physics.plasm-ph↗