HawkesLLM: Semantic Uncertainty Propagation in Agentic Text Simulation

2026-05-21

AI Summary

This work addresses the path dependency problem in agent-based textual simulation caused by early-stage semantic uncertainty. To this end, the authors propose HawkesLLM, a novel framework that introduces multivariate Hawkes processes into this domain for the first time, explicitly modeling the temporal activation and influence dynamics among agents while decoupling temporal influence from text generation. By incorporating a compact prompt memory mechanism, HawkesLLM effectively disentangles local and global semantic drift, significantly enhancing semantic alignment between later-generation outputs and local reference texts under constrained memory budgets. The approach is evaluated on the GDELT news cascade dataset, demonstrating its effectiveness in event-level text generation.

Abstract

Agentic text-simulation systems write in sequence, with each item becoming possible context for later steps. That makes uncertainty path-dependent: an early ambiguity can affect later outputs. This paper studies this problem with HawkesLLM, a framework that separates temporal influence modeling from text generation. We represent the cascade as a network whose nodes are text-generating agents. A multivariate Hawkes process models how these nodes activate over time and which earlier node outputs should influence later prompts. A language model then writes each new event from the compact memory selected by this temporal model. We evaluate the framework on a held-out Global Database of Events, Language, and Tone (GDELT) news-cascade case study. The diagnostics track semantic alignment with local held-out references and separate local drift from global drift. In this setting, HawkesLLM improves late-stage semantic alignment under a compact prompt-memory budget.

Problem

• semantic uncertainty
• agentic text simulation
• uncertainty propagation
• path dependence
• text generation

Innovation

• Hawkes process
• semantic uncertainty
• agentic simulation
• temporal influence modeling
• prompt memory