Reference library
Recurrent Memory Transformer
Paper summary
Recurrent Memory Transformer adds memory tokens passed between sequence segments. Training teaches these tokens to carry compact information, allowing segment recurrence without redesigning the core Transformer.
Relevance to the project
A comparison for explicit bounded history versus compressed persistent state. Test what information survives beyond the accessible history window.
Compact memory can lose information; a memory token is not unlimited storage.
Source record
- Reference ID
- recurrent-memory-transformer
- Archived version
- 2207.06881v2
- Summary provenance
- reference-library/papers/recurrent-memory-transformer/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.