Human-Like Lifelong Memory: A Neuroscience-Grounded Architecture for Infinite Interaction
Diego C. Lerma-Torres · 2026
Read Human-Like Lifelong Memory: A Neuroscience-Grounded Architecture for Infinite Interaction by Diego C. Lerma-Torres (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/arxiv-2603-29023.txt.
Description
Lerma-Torres proposes a neuroscience grounded memory architecture for long term interaction with language models. Bigger context windows alone can degrade reasoning, even with perfect retrieval. The design draws on complementary learning systems theory, belief hierarchies, dual process cognition, and fuzzy-trace theory. Memory carries valence, not just content, through pre-computed emotion summaries in a belief hierarchy. Retrieval defaults to fast automatic activation, with planned search only when needed. A thalamic gateway tags and routes information between stores. Over time, the system shifts toward fast processing, like expert skill.
Catalog record
| CREATOR | Diego C. Lerma-Torres |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41523 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.29023