Key-value memory in the brain
Open accessAbstract
Classical models of memory in psychology and neuroscience rely on similarity-based retrieval of stored patterns, where similarity is a function of retrieval cues and the stored patterns. While parsimonious, these models do not allow distinct representations for storage and retrieval, despite their distinct computational demands. Key-value memory systems, in contrast, distinguish representations used for storage (values) and those used for retrieval (keys). This allows key-value memory systems to optimize simultaneously for fidelity in storage and discriminability in retrieval. We review the computational foundations of key-value memory, its role in modern machine learning systems, related ideas from psychology and neuroscience, applications to a number of empirical puzzles, and possible biological implementations.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Human memory types and working mechanisms can inspire the development of more scientific and reasonable memory-enhanced AI systems.
“The different types of human memory and their working mechanisms can greatly inspire us to develop more scientific and reasonable memory-enhanced AI systems [13-16].”
From From Human Memory to AI Memory: A Survey on Memory Mechanisms in the Era of LLMs (2025) · cites this paper as [15]
Cites
Rethinking Attention with Performers
Choromanski, K., Likhosherstov, V., Dohan, D., Song, X., Gane, A., Sarlos, T., Hawkins, P., Davis, J., Mohiuddin, A., Kaiser, L., et al. (2020) · arXiv · doi:10.48550/arxiv.2009.14794 · Cited by 695 · cited 1× · see citing works
“Several studies 29,30 have shown how to rigorously approximate softmax attention using random features.”
Specific evidence of low-dimensional continuous attractor dynamics in grid cells
Yoon, K., Buice, M. A., Barry, C., Hayman, R., Burgess, N., & Fiete, I. R. (2013) · Nature Neuroscience · doi:10.1038/nn.3450 · Cited by 296 · cited 1× · see citing works
“Consisent with the non-learned assumption, grid cells exhibit a fixed relational organization regardless of environment and behavioral state 57–60 .”
Self-citationIrie, K., Csordás, R., & Schmidhuber, J. (2022) · arXiv · doi:10.48550/arxiv.2202.05798 · Cited by 6 · cited 3× · see citing works
“Generalizing a classic result on kernel machines 41 , Irie and colleagues 12 showed that this construction is equivalent (in the sense of producing the same output for a given input) to the following linear key-value memory: where (using our earlier notation) α = qK ⊤ with v n = e n , k n = x n ,…”
Self-citation
Flagged first, then by citations
Cited by
Sorted by citations
From Human Memory to AI Memory: A Survey on Memory Mechanisms in the Era of LLMs
Wu, Y., Liang, S., Zhang, C., Wang, Y., Zhang, Y., Guo, H., Tang, R., & Liu, Y. (2025) · arXiv · doi:10.48550/arxiv.2504.15965 · Cited by 3 · see citing works
Flexible Prefrontal Control over Hippocampal Episodic Memory for Goal-Directed Generalization
Zheng, Y., Wolf, N., Ranganath, C., O'Reilly, R. C., & McKee, K. L. (2025) · Proceedings of Cognitive Computational Neuroscience 2025 · doi:10.32470/jnqbpro
- Authors
- Gershman, S. J., Fiete, I., & Irie, K.
- Affiliations
- Harvard University (US)
- McGovern Institute for Brain Research (US)
- Corresponding author
- Gershman, Samuel J.
- Venue
- arXiv
- Publisher
- arXiv
- Published
- 2025
- Volume
- 113
- Issue
- 11
- Pages
- 1694-1707.e1
- ISSN
- 0896-6273 (ISSN-L)
- arXiv
- 2501.02950
- OpenAlex
- W4408896211
- Cited by
- 21
- References
- 3
- Citations by year
- 2026: 14 · 2025: 7
- FWCI
- 3.50
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- Neuron (acceptedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by)
- arXiv (Cornell University) (cc-by)
- Funders
- U.S. Department of Defense
- Multidisciplinary University Research Initiative (W911NF-23-1-0277)
- Army Research Office (W911NF-23-1-0277, W911NF)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › Memory and Neural Mechanisms
- Life Sciences › Neuroscience › Cognitive Neuroscience › Memory Processes and Influences
- Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
- Keywords
- key-value memory, memory retrieval, similarity search, memory fidelity, memory models
- MeSH terms
- Machine Learning, Animals, Brain, Humans, Memory, Models, Neurological, Mental Recall
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
