返回首页
arXiv AI··论文与技术

Rethinking User Encoders: Routing Diffusion States for Scenario-Adaptive User Representation

中文摘要

现有LLM用户编码器生成静态表示,限制双向信息与跨模态交互。新研究提出扩散状态路由,实现场景自适应用户表示,提升适应性。

English Summary

Existing LLM user encoders yield static representations, limiting bidirectional flow and cross-modal interactions. This paper introduces routing diffusion states for scenario-adaptive user representation, improving adaptability.

原文节选

arXiv:2610.22400v1 Announce Type: new Abstract: Existing large language model (LLM)-based user encoders serialize heterogeneous behavioral logs and compress them into static representations through autoregressive (AR) prediction. However, their unidirectional constraint limits bidirectional information flow, hindering cross-modal interactions and the joint preservation of multi-source signals. Consequently, the learned static representations struggle to accommodate the distinct evidence requirements of diverse downstream scenarios. We address these limitations by unlocking the iterative denoising process of diffusion language models (dLLMs) as a multi-granularity reconstruction space. Based on this insight, we propose FINDERS (FactorIzatioN of Diffusion statEs for Routing across Scenarios), a router-equipped dLLM framework comprising four operations: denoise, factorize, route, and compress. Learnable meta-slots organize complementary denoising states into transferable cross-modal semantics and modality-specific behavioral factors. To provide grounded supervision, we construct BehaviorQA, a large-scale corpus containing explicit behavioral rationales. Guided by scenario queries, a l…