HawkesLLM: Semantic Uncertainty Propagation in Agentic Text Simulation
Quick Answer
HawkesLLM introduces a framework for agentic text simulation that models temporal influence using a multivariate Hawkes process, enhancing late-stage semantic alignment in text generation.
Quick Take
Evaluated on the GDELT news-cascade dataset, it effectively separates local drift from global drift while maintaining a compact prompt-memory budget.
Key Points
- HawkesLLM separates temporal influence modeling from text generation.
- Utilizes a multivariate Hawkes process for node activation over time.
- Improves late-stage semantic alignment under compact memory constraints.
- Evaluated using the GDELT news-cascade case study.
- Tracks semantic alignment with local held-out references.
Paper Resources
Article Excerpt
From source RSS / original summaryarXiv: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.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →TriAgent: Divergence-Aware Committees for Cost-Efficient Financial Sentiment Analysis
TriAgent introduces a cost-efficient multi-agent system for financial sentiment analysis, combining VADER, FinBERT, and Qwen2.5. It achieves an F1 score of ~0.87 with significant savings of $9.3M/year at a 10M-user scale compared to GPT-4o-mini, while also detecting hallucinations with an AUC of 0.90.