Research

Paper

AI LLM March 17, 2026

Trained Persistent Memory for Frozen Encoder--Decoder LLMs: Six Architectural Methods

Authors

Hong Jeong

Abstract

Frozen encoder--decoder language models are stateless: the latent representation is discarded after every forward pass, so no information persists across sessions. This paper presents a \textbf{proof-of-concept pilot study} showing that persistent memory in the \emph{continuous latent space} of a frozen LLM is feasible -- even under severe resource constraints (a single frozen Flan-T5-XL backbone, small trainable adapters, a single dataset). We implement six architectural methods spanning three injection points and four write mechanisms; unlike text-level memory systems, every write and read is a differentiable operation on dense vectors. After training only the adapter, the memory bank continues to accumulate at inference time without gradients, enabling \emph{conversational learning}. Under a forgetting-curve evaluation on LoCoMo at two capacity scales (1$\times$ and 10$\times$), the stateless baseline scores exactly zero; at 10$\times$ all six trained adapters produce positive memory-recall curves; at 1$\times$ three methods collapse, revealing capacity as a critical design parameter. Because the memory bank is a compact numerical array, it can be scaled to arbitrarily large capacity without altering the backbone. We argue that full end-to-end training with larger models, larger data, and orders-of-magnitude larger memory will yield substantially stronger results; this pilot study establishes the feasibility baseline and design-space taxonomy that such efforts require.

Metadata

arXiv ID: 2603.16413
Provider: ARXIV
Primary Category: cs.LG
Published: 2026-03-17
Fetched: 2026-03-18 06:02

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.16413v1</id>\n    <title>Trained Persistent Memory for Frozen Encoder--Decoder LLMs: Six Architectural Methods</title>\n    <updated>2026-03-17T11:51:21Z</updated>\n    <link href='https://arxiv.org/abs/2603.16413v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.16413v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Frozen encoder--decoder language models are stateless: the latent representation is discarded after every forward pass, so no information persists across sessions. This paper presents a \\textbf{proof-of-concept pilot study} showing that persistent memory in the \\emph{continuous latent space} of a frozen LLM is feasible -- even under severe resource constraints (a single frozen Flan-T5-XL backbone, small trainable adapters, a single dataset). We implement six architectural methods spanning three injection points and four write mechanisms; unlike text-level memory systems, every write and read is a differentiable operation on dense vectors. After training only the adapter, the memory bank continues to accumulate at inference time without gradients, enabling \\emph{conversational learning}. Under a forgetting-curve evaluation on LoCoMo at two capacity scales (1$\\times$ and 10$\\times$), the stateless baseline scores exactly zero; at 10$\\times$ all six trained adapters produce positive memory-recall curves; at 1$\\times$ three methods collapse, revealing capacity as a critical design parameter. Because the memory bank is a compact numerical array, it can be scaled to arbitrarily large capacity without altering the backbone. We argue that full end-to-end training with larger models, larger data, and orders-of-magnitude larger memory will yield substantially stronger results; this pilot study establishes the feasibility baseline and design-space taxonomy that such efforts require.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-03-17T11:51:21Z</published>\n    <arxiv:primary_category term='cs.LG'/>\n    <author>\n      <name>Hong Jeong</name>\n    </author>\n  </entry>"
}