Intrinsic Entropy of Context Length Scaling in LLMs

ICLROral2026

Authors
Jingzhe Shi, Qinwei Ma, Hongyi Liu, Hang Zhao, Jenq-Neng Hwang, Lei Li
Affiliation
Tsinghua University
Venue
ICLR 2026
Track
Oral

TL;DR

We propose to use Intrinsic Entropy for understanding impact of context length on Language Modeling, and conduct experiments to validate theoretical assumptions and deductions with language and synthetic datasets.

Opening excerpt from the authors’ abstract. source

Read the paper

Topics

language model dataset llm

← All ICLR 2026 Oral papers · Browse the whole archive