Philippe Laban
2 papers · Latest:
Natural Language Processing
Measuring and Mitigating the Distributional Gap Between Real and Simulated User Behaviors
This paper introduces a method to measure the distributional gap between real and simulated user behaviors, evaluating 24 LLM-based simulators.
2605.07847
Natural Language ProcessingLLMs Corrupt Your Documents When You Delegate
LLMs silently corrupt documents by introducing severe errors during long delegated workflows, degrading content by an average of 25% in frontier models.
2604.15597
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.