@yoheinakajima: your hippocampus fast-captures events (logs) that are computed on to hold a current state, and through selective replay…
Summary
The tweet explains how the hippocampus fast-captures events to hold a current state and selectively replays to update the cortex model, feeding new priors back into computations.
Similar Articles
@yoheinakajima: https://x.com/yoheinakajima/status/2081741659260477666
This thread explores how the brain's dual memory systems (hippocampus and neocortex) offer lessons for building long-running AI agents, arguing that agents need a fast episodic capture and slow consolidation mechanism to avoid catastrophic interference, rather than relying solely on frozen models with temporary scaffolding.
History Replays Itself: Unleashing Long-Context Capability in Linear Attention
The paper introduces NoteBank, a training-free sparse episodic memory that augments linear attention models by archiving selected text and replaying it through native computation, boosting Gated DeltaNet's long-context retrieval accuracy on RULER from 1.7% to 70.3% at one million tokens with negligible FLOPs overhead.
@aakashgupta: Anthropic just shipped sleep into agents. When you sleep, your hippocampus replays the day's neural sequences to the co…
Anthropic has introduced a new 'sleep' mechanism for AI agents inspired by biological hippocampal replay and dreaming to extract patterns and reorganize memories, aiming to prevent capability plateaus associated with raw context window reliance.
@omarsar0: NEW paper worth reading. (bookmark it) The basic idea is to pair a compressive recurrent state with a small exact memor…
HOLA (Hippocampal Linear Attention) augments linear attention with a bounded exact KV cache inspired by hippocampal memory, improving long-range recall and perplexity without sacrificing efficiency. At 340M parameters, it outperforms full-attention Transformers on Wikitext and achieves robust needle recall up to 32k tokens.
@yoheinakajima: check out http://activegraph.ai, flattening all of it into a single event log :)
ActiveGraph provides a persistent, replayable world for long-running agents using an append-only event log, enabling fork, diff, and lineage capabilities.