Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

Transformer memory

Transformer memory refers to the internal capacity of a Transformer neural network to store, organize, and recall knowledge directly within its learned model parameters rather than relying on external databases or indexing structures. In this paradigm, information such as document content, associations, and factual data is encoded across the weights of the network during training. This parameter-based storage enables the model to function as a self-contained information retrieval and generation system, allowing it to process input queries and map them directly to relevant document identifiers, answers, or representations through end-to-end differentiable operations.

1 item

Transformer Memory as a Differentiable Search Index

Transformer Memory as a Differentiable Search Index

Yi Tay, Vinh Tran, Mostafa Dehghani, Jianmo Ni, Dara Bahri, Harsh Mehta, Zhen Qin, Kai Hui, Zhe Zhao, Jai Prakash Gupta, Tal Schuster, William W. Cohen, Donald Metzler

OrganizationsGoogle

Why you should read this

Proposes the Differentiable Search Index, demonstrating that a single Transformer can internalize an entire text corpus within its parameters to map queries directly to document identifiers, outperforming standard dual-encoder retrieval systems.

In this paper, we demonstrate that information retrieval can be accomplished with a single Transformer, in which all information about the corpus is encoded in the parameters of the model. To this end, we introduce the Differentiable Search Index (DSI), a new paradigm that learns a text-to-text model that maps string queries directly to relevant docids; in other words, a DSI model answers queries directly using only its parameters, dramatically simplifying the whole retrieval process. We study variations in how documents and their identifiers are represented, variations in training procedures, and the interplay between models and corpus sizes. Experiments demonstrate that given appropriate design choices, DSI significantly outperforms strong baselines such as dual encoder models. Moreover, DSI demonstrates strong generalization capabilities, outperforming a BM25 baseline in a zero-shot setup.

Added

2026-09-26