Paper Detail

KITE: Retrieval as a Decision Procedure for Long-Term Conversational Memory

Feng Li, Kailun Tan, Zhibo Zhou, Youxiang Wang, Lejun Liu, Huaigu Li, Fei Du, Huilong An

arxiv Score 12.3

Published 2026-10-30 · First seen 2026-11-02

Research Track A · General AI

Abstract

Many long-term memory systems for LLM agents are built around embedding-based retrieval. Al- though effective for semantic matching, similarity ranking alone does not establish whether re- trieved memories satisfy the conditions implied by a question. In this paper, we introduce KITE (Knowledge-Indexed Temporal Evidence), a vector-free framework that formulates memory re- trieval as an explicit decision procedure. KITE organizes multi-session conversations as structured knowledge indexed by a governed topic vocabulary, which evolves through controlled expansion and consolidation. At query time, questions are compiled into inspectable retrieval plans over the same knowledge index that guides memory construction. The plans are then executed through fixed lookup and ranking operations, while sparse lexical retrieval recovers surface details omitted from the structured representation. Experiments on LoCoMo and LongMemEval-S show that KITE achieves 93.5% and 86.0% accuracy while using only about 1.6k reader-context tokens per ques- tion. These results demonstrate that conversational memory can move beyond similarity ranking by unifying memory construction and query compilation around a governed index, turning retrieval into an inspectable decision procedure for determining which evidence satisfies the question.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{li2026kite,
  title = {KITE: Retrieval as a Decision Procedure for Long-Term Conversational Memory},
  author = {Feng Li and Kailun Tan and Zhibo Zhou and Youxiang Wang and Lejun Liu and Huaigu Li and Fei Du and Huilong An},
  year = {2026},
  abstract = {Many long-term memory systems for LLM agents are built around embedding-based retrieval. Al- though effective for semantic matching, similarity ranking alone does not establish whether re- trieved memories satisfy the conditions implied by a question. In this paper, we introduce KITE (Knowledge-Indexed Temporal Evidence), a vector-free framework that formulates memory re- trieval as an explicit decision procedure. KITE organizes multi-session conversations as structured knowledge indexed by a go},
  url = {https://arxiv.org/abs/2610.41521},
  keywords = {cs.IR},
  eprint = {2610.41521},
  archiveprefix = {arXiv},
}

Metadata

{}