让大模型持续学习用户偏好，避免遗忘旧兴趣。
SPRInG: Continual LLM Personalization via Selective Parametric Adaptation and Retrieval-Interpolated Generation
- 只更新有明显偏好转折的对话，避免被无关内容干扰。
- 在长文本生成任务中，性能超越现有方法显著。
- 适合需要长期记住用户喜好的智能助手场景。
个性化大语言模型通常依赖静态检索或一次性适配，假设用户偏好恒定不变。然而真实交互中用户兴趣持续演变，对模型适应偏好漂移又不遗忘旧偏好提出挑战。标准持续学习方法常因在嘈杂交互流中盲目更新而失效。为此，我们提出SPRInG，一种半参数化框架，实现有效持续个性化。训练时采用基于似然的评分函数识别高新颖性交互，仅对偏好漂移信号选择性更新用户适配器，同时将难学残差保存至重放缓冲区。推理时通过严格相关性过滤，并以概率插值融合参数知识与历史检索结果。在长文本个性化生成基准测试中，SPRInG显著优于现有基线，验证其在真实场景下持续个性化的鲁棒性。
原文摘要 · Abstract (English)
Personalizing Large Language Models typically relies on static retrieval or one-time adaptation, assuming user preferences remain invariant over time. However, real-world interactions are dynamic, where user interests continuously evolve, posing a challenge for models to adapt to preference drift without catastrophic forgetting. Standard continual learning approaches often struggle in this context, as they indiscriminately update on noisy interaction streams, failing to distinguish genuine preference shifts from transient contexts. To address this, we introduce SPRInG, a novel semi-parametric framework designed for effective continual personalization. During training, SPRInG employs drift-driven selective adaptation, which utilizes a likelihood-based scoring function to identify high-novelty interactions, selectively updating the user-specific adapter on drift signals while preserving hard-to-learn residuals in a replay buffer. During inference, we apply strict relevance gating and fuse parametric knowledge with retrieved history via probability interpolation. Experiments on the long-form personalized generation benchmark demonstrate that SPRInG significantly outperforms existing baselines, validating its robustness for real-world continual personalization.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。