pen Life Lab
Back to reference library

Reference library

Titans: Learning to Memorize at Test Time

Paper summary

Titans adds a neural long-term memory that updates at test time alongside shorter-term context processing. Surprise-sensitive memory updates and architectural variants are evaluated for sequence tasks and long-context use.

Relevance to the project

A close comparison for explicit histories plus mutable learned memory. Distinguish state, fast weights, readout, and the origin of an update signal.

Language modelsMemorySelf-modifying networksLong-context processing

The memory rule and external training objective are supplied rather than evolved within an environment.

Source record

Reference ID
arxiv-2501-00663
Archived version
2501.00663v1
Summary provenance
reference-library/papers/arxiv-2501-00663/metadata.json · summary
Review status
initial orientation summary; not a critical full-paper review
Reading scope
Selected abstract/opening and available closing discussion/conclusion passages in the preserved text extract; broader project notes provide context. No reproduction of experiments, exhaustive methods review, or equation/figure audit.