CARD: Cluster-level Adaptation with Reward-guided Decoding for Personalized Text Generation

类型：arxiv
标识：2601.06352
链接：https://arxiv.org/abs/2601.06352
主分类：engineering
形态：application
TLDR：Adapting large language models to individual users remains challenging due to the tension between fine-grained personalization and scalable deployment. We present CARD, a hierarchical framework that achieves effective personalization through progressive refinement. CARD first clusters users according to shared stylistic patterns and learns group-specific LoRA adapters, enabling robust generalization and strong low-resource performance. To capture individual differences within each cluster, we propose an implicit preference learning mechanism that contrasts user-authored text with cluster-level
待LLM分类：否
来源文件：
/inbox/tom/_candidates/2026-09-28-agent-rag-longcontext-candidates.json

2026-08-09

仓库 griver/Q-RAG

2026-08-13

笔记精读与批判 v2 · Image Tokenizers as Visual Languages in Unified Multimodal Models · 2026-09-12 中午棒（重写覆盖）

2026-09-12

笔记 Cool Papers · cs.CL (papers.cool) · RSS 摘要

2026-08-16

笔记精读：Scaling Beyond Context — 多模态 RAG for Document Understanding Survey

2026-09-14

笔记 2026-09-05 精读批判 | 多模态长文档 RAG 与高效 VLM · v2 覆盖版

2026-09-05

arXiv

2601.06352

TLDR

Adapting large language models to individual users remains challenging due to the tension between fine-grained personalization and scalable deployment. We present CARD, a hierarchical framework that achieves effective personalization through progressive refinement. CARD first clusters users according to shared stylistic patterns and learns group-specific LoRA adapters, enabling robust generalization and strong low-resource performance. To capture individual differences within each cluster, we propose an implicit preference learning mechanism that contrasts user-authored text with cluster-level

CARD: Cluster-level Adaptation with Reward-guided Decoding for Personalized Text Generation

相关 · 按语义匹配