SOURCE-LINKED INTELLIGENCE

LatentPress: Context Compression Beyond Text and Vision

arXiv · AI, language, vision and robotics · article · Sep 1, 2026 · UTC

Compressed context is usually carried as human-readable text or as rendered images that must be decoded, even when its consumer is a language model. We introduce LatentPress, which writes conversational histories and long documents into a third representation: continuous memory tokens that a frozen decoder reads directly through its input-embedding interface, with no text reconstruction at inference. A small reader-matched writer compresses $4$-$16\times$ while training only an adapter (4.2M-26.2M parameters, $\sim\!0.1\%$ of the decoder). On LongMemEval, LatentPress reaches $0.504$ accuracy a

Read original source ↗ Open in workspace

recordType

paper

region

Global