HOLA: Adding Hippocampus-like Precise Memory to Linear Attention
omarsar0 · x · 2026-07-03
A new paper proposes HOLA: adding a bounded exact KV memory module on top of the compressed recurrent states of linear attention or state space models, acting as a "hippocampal" supplement. Linear attention compresses the entire context into a fixed-size state to achieve O(1) memory, but early facts get overwritten and needle recall degrades when numerous key-value associations compete. HOLA restores long-term memory capabilities without sacrificing the efficiency of linear attention.
More from Research
- HarnessMM cuts agent tokens per task by up to 88% while raising success rates — omarsar0 · 2026-10-26
- Ben Recht: The world needs a truly open language model, and it's easier to build than you think — beenwrekt · 2026-10-26
- New Blog Breaks Black-Box Distillation Into Five Practical Steps — chrisrohlf · 2026-10-26
- DetectorScope turns black-box AI text detectors into auditable ones with sparse autoencoders — _akhaliq · 2026-10-26
- OPST paper shows LLMs self-improve with zero human data, guided only by topic descriptions — _akhaliq · 2026-10-26
- Tencent's HiSTEVE 1.0: hierarchical VLA agent plays Minecraft with native keyboard and mouse, hits 85% SOTA — _akhaliq · 2026-10-26