Papers
arxiv:2407.01437

Needle in the Haystack for Memory Based Large Language Models

Published on Jul 1, 2024
Authors:
,
,
,
,

Abstract

Larimar, a language model with an external associative memory, outperforms larger transformer-based models on long-context recall tasks without task-specific training.

AI-generated summary

Current large language models (LLMs) often perform poorly on simple fact retrieval tasks. Here we investigate if coupling a dynamically adaptable external memory to a LLM can alleviate this problem. For this purpose, we test Larimar, a recently proposed language model architecture which uses an external associative memory, on long-context recall tasks including passkey and needle-in-the-haystack tests. We demonstrate that the external memory of Larimar, which allows fast write and read of an episode of text samples, can be used at test time to handle contexts much longer than those seen during training. We further show that the latent readouts from the memory (to which long contexts are written) control the decoder towards generating correct outputs, with the memory stored off of the GPU. Compared to existing transformer-based LLM architectures for long-context recall tasks that use larger parameter counts or modified attention mechanisms, a relatively smaller size Larimar is able to maintain strong performance without any task-specific training or training on longer contexts.

Community

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2407.01437 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2407.01437 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2407.01437 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.