Selective Memory for Artificial Intelligence: Write-Time Gating with Hierarchical Archiving

TL;DR

This paper introduces write-time gating with hierarchical archiving for AI memory, selectively filtering knowledge based on salience to improve accuracy and robustness, outperforming traditional retrieval-augmented methods especially under distractor conditions.

Contribution

The paper presents a novel write-time gating mechanism that filters incoming knowledge objects using composite salience scores, maintaining version chains and outperforming read-time filtering methods in accuracy and efficiency.

Findings

Write gating achieves 100% accuracy versus 13% for ungated stores.

Under high distractor ratios, write gating maintains 100% accuracy while read-time filtering collapses.

Validation across multiple datasets confirms the structural advantage of write gating.

Abstract

Retrieval-augmented generation stores all content indiscriminately, degrading accuracy as noise accumulates. Parametric approaches compress knowledge into weights, precluding selective updates. Neither mirrors biological memory, which gates encoding based on salience and archives rather than deletes superseded information. We introduce write-time gating that filters incoming knowledge objects using composite salience scores (source reputation, novelty, reliability) while maintaining version chains that preserve prior states. Using real LLM evaluation without oracle access to quality labels, write gating achieves 100 percent accuracy versus 13 percent for ungated stores. The critical finding emerges under distractor scaling: at 8:1 distractor ratios, read-time filtering (Self-RAG) collapses to 0 percent while write gating maintains 100 percent, revealing a structural advantage of write-time over read-time curation. Validation on Wikipedia (20 entities), procedurally generated pharmacology data, and 2026 arXiv papers confirms these findings. The gating advantage scales inversely with parametric memory support: +25pp for Wikipedia, +48pp for post-cutoff arXiv, +65pp for procedural data with zero training knowledge. Signal ablation confirms the method does not depend on oracle-correlated metadata. Write gating matches Self-RAG accuracy at one-ninth the query-time cost.