Hippocampo-neocortical interaction as compressive retrieval-augmented generation
Eleanor Spens, Neil Burgess · 2026
Read Hippocampo-neocortical interaction as compressive retrieval-augmented generation by Eleanor Spens, Neil Burgess (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13447824.txt.
Description
Eleanor Spens and Neil Burgess model the exchange between episodic and semantic memory as packed retrieval-augmented generation in 2026 work. Episodic memory in the hippocampus holds single events, while semantic memory in the neocortex holds general knowledge, and their cooperation underpins recall and problem solving. How the two systems trade information over time has lacked a concrete account.
Their model stores experiences in the hippocampus in compressed form, then replays them to train a neocortical network. The network learns the gist of episodes and patterns that carry over to new problems. During recall, hippocampal retrieval feeds fitting episodes into working memory, where neocortical knowledge shapes the rebuild. Model runs reproduce known effects such as schema-driven warping of recall, showing how detail slowly gives way to gist.
Catalog record
| CREATOR | Eleanor Spens, Neil Burgess |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 109666 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |