SOURCE-LINKED INTELLIGENCE

Dynamic Semantic Compression for Efficient Latent-Space Inference in Large Language Models

arXiv · AI, language, vision and robotics · article · Sep 14, 2026 · UTC

Large Language Models (LLMs) primarily perform inference at the token level, resulting in substantial memory overhead and compromised computational efficiency. In this paper, we propose a Dynamic Semantic Extraction and Inference (DSEI) framework, which achieves segment-level inference within the latent space through a two-stage training strategy. First, we construct a Dynamic Semantic Autoencoder (DSAE) via self-supervised learning. DSAE dynamically extracts segment-level semantics and compresses them into compact latent representations via adaptive semantic weighting and gated fusion. Subseq

Read original source ↗ Open in workspace

recordType

paper

region

Global