Saved in:
Bibliographic Details
Main Authors: Den, Yu-Chen, Shih, Yung-Yu, Tam, Zhi Rui, Chen, Kuan-Yu, Cheng, Pu-Jen, Chen, Yun-Nung, Yang, Eugene
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.26902
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Generative retrieval (GR) maps queries directly to document identifiers (docids) using parametric knowledge, However, this design makes corpus expansion costly: adding new documents requires updating model parameters to encode new document-docid associations incurs repeated training and catastrophic forgetting of previously indexed documents. In this work, we revisit incremental GR as an in-context retrieval problem, where newly added documents are supplied as inference-time document-docid evidence. We propose ICICLE, an in-context indexing framework that performs source-aware docid generation over both parametric memory and context-provided document-docid pairs. ICICLE combines a `[COPY]`-based routing mechanism, preference-based calibration, and large context adaptation to distinguish context-grounded retrieval from parametric retrieval. Experiments on MS MARCO and NQ320K show that ICICLE improves retrieval of newly introduced documents while preserving seen-document retention without corpus-specific retraining. Our analysis further shows that high-shot degradation is mainly caused by routing failure, highlighting source-selection calibration as a key bottleneck for scaling in-context generative retrieval.