# Preprint proposes near-constant-memory long-context recall method

_Monday, September 14, 2026 at 12:00 AM EDT · AI, Science · Latest · Tier 2 — Notable_

A preprint proposes reconstructing document facts from residual vectors in a language model's feed-forward-layer activations, rather than keeping the original document in context. The authors say the method maintains near-constant GPU memory use as context length rises and requires no additional training or weight fine-tuning. In their experiments, they report answering single-fact questions in two-million-token story contexts where prior methods failed. The paper frames the approach as a response to memory usage that otherwise grows with input length.

## Sources

- [cs.CL updates on arXiv.org](https://arxiv.org/abs/2609.12686)

---
Canonical: https://techandbusiness.org/newswire/3alz6djI75LMpPFu14-lfx
Retrieved: 2026-09-14T17:28:27.002Z
Publisher: Tech & Business (techandbusiness.org)