隐性偏见如何在LLM长期记忆中积累和传播
How Implicit Bias Accumulates and Propagates in LLM Long-term Memory
AI总结：
本文研究了LLM长期记忆中隐性偏见的积累与传播，提出动态内存标记方法以缓解偏见问题。
摘要：
长期记忆机制使大型语言模型（LLMs）能够在长时间的交互生命周期中保持连续性和个性化，但它们也引入了与公平性相关的新型和未被深入研究的风险。在本工作中，我们研究了隐性偏见，即微妙的统计偏见，在配备了长期记忆的LLMs中如何积累和传播。为了支持系统分析，我们引入了基于决策的隐性偏见（DIB）基准，这是一个包含3776个决策场景的大型数据集，覆盖九个社会领域，旨在量化长期决策过程中的隐性偏见。使用一个现实的长周期模拟框架，我们评估了六个最先进的LLMs，结合三种代表性的记忆架构，在DIB上的表现，并证明LLMs的隐性偏见并非静态，而是随时间推移而加剧，并在无关领域之间传播。我们进一步分析了缓解策略，并展示了静态系统级提示基线提供有限且短暂的去偏效应。为了解决这一限制，我们提出了动态内存标记（DMT），一种代理干预，它在内存写入时强制实施公平性约束。广泛实验结果表明，DMT显著减少了偏见积累，并有效遏制了跨领域偏见传播。
英文摘要：
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.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
