Saved in:
Bibliographic Details
Main Authors: Li, Yangning, Chen, Shaoshen, Li, Yinghui, Chen, Yankai, Zheng, Hai-Tao, Wang, Hui, Jiang, Wenhao, Yu, Philip S.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.04550
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • The quadratic complexity of self-attention constrains Large Language Models (LLMs) in processing long contexts, a capability essential for many advanced applications. Context compression aims to alleviate this computational bottleneck while retaining critical semantic information. However, existing approaches often fall short: explicit methods may compromise local detail, whereas implicit methods can suffer from positional biases, information degradation, or an inability to capture long-range semantic dependencies. We propose AdmTree, a novel framework for adaptive, hierarchical context compression with a central focus on preserving high semantic fidelity while maintaining efficiency. AdmTree dynamically segments input based on information density, utilizing gist tokens to summarize variable-length segments as the leaves of a semantic binary tree. This structure, together with a lightweight aggregation mechanism and a frozen backbone LLM (thereby minimizing new trainable parameters), enables efficient hierarchical abstraction of the context. By preserving fine-grained details alongside global semantic coherence, mitigating positional bias, and dynamically adapting to content, AdmTree robustly retains the semantic information of long contexts.