HawkesLLM: Semantic Uncertainty Propagation in Agentic Text Simulation

Researchers introduced HawkesLLM, a framework that models how uncertainty propagates through sequential text generation in agentic simulation systems. The framework uses a multivariate Hawkes process to track temporal influences between text-generating agents, then applies a language model to write new events based on selected prior outputs. In tests using GDELT news-cascade data, HawkesLLM improved late-stage semantic alignment while operating under a compact prompt-memory budget.

arXiv:2605.23043v1 Announce Type: new 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.