Coupled Spiking Memory for Budgeted Long-Dialogue Question Answering with Small Language Models

Longitudinal conversational assistance requires access to information that is scattered across sessions and changes over time. A small language model can retrieve this information from an external log, but independently ranking each record ignores the sequence in which the record was produced. We present Coupled Spiking Memory (CSM), an evidence-supervised log controller that augments each record with a causal temporal trace before query-dependent retrieval. CSM adapts two-compartment spiking dynamics, reads the post-reset somatic voltage, and learns a residual correction to lexical relevance while leaving the language model frozen. Whole-record selection restricts the retrieved context to a fixed token budget and preserves chronological presentation. With a frozen 3B reader, CSM obtains 30.33 mean QA F1 on 192 EvoEmo questions across five training seeds, exceeding BM25, BGE-M3, and a matched GRU controller by 0.31, 1.28, and 0.76 points, respectively. It uses 123,907 trainable controller parameters. Frozen transfer to LoCoMo and a three-budget analysis reveal that the benefit depends on the readout, evidence budget, and balance between answering and abstention. These results establish a compact spiking-controller approach to conversational evidence retrieval and identify the conditions that limit its transfer.

Authors

Institutions

Publication Details

Journal
Qeios
Published
2026-11-17
DOI
https://doi.org/10.32388/89xyao
Primary Topic
Information Retrieval and Search Behavior
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Coupled Spiking Memory for Budgeted Long-Dialogue Question Answering with Small Language Models

Jiahao Qin, Yiwen Wang
2 links
Qeios
Information Retrieval and Search Behavior
preprint

Coupled Spiking Memory for Budgeted Long-Dialogue Question Answering with Small Language Models

Jiahao Qin, Yiwen Wang
preprint en

Abstract

Longitudinal conversational assistance requires access to information that is scattered across sessions and changes over time. A small language model can retrieve this information from an external log, but independently ranking each record ignores the sequence in which the record was produced. We present Coupled Spiking Memory (CSM), an evidence-supervised log controller that augments each record with a causal temporal trace before query-dependent retrieval. CSM adapts two-compartment spiking dynamics, reads the post-reset somatic voltage, and learns a residual correction to lexical relevance while leaving the language model frozen. Whole-record selection restricts the retrieved context to a fixed token budget and preserves chronological presentation. With a frozen 3B reader, CSM obtains 30.33 mean QA F1 on 192 EvoEmo questions across five training seeds, exceeding BM25, BGE-M3, and a matched GRU controller by 0.31, 1.28, and 0.76 points, respectively. It uses 123,907 trainable controller parameters. Frozen transfer to LoCoMo and a three-budget analysis reveal that the benefit depends on the readout, evidence budget, and balance between answering and abstention. These results establish a compact spiking-controller approach to conversational evidence retrieval and identify the conditions that limit its transfer.

Qeios
University of Liverpool (GB), Fudan University (CN)
Information Retrieval and Search Behavior
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon