Tag
The paper introduces the Normalized Context Utilization (NCU) metric to quantify contextual information gain in RAG systems, challenging scaling laws by showing that small language models can match or outperform larger models in strict factual extraction due to lower 'Prior Dominance', and that a commercial API frequently overrode external evidence in adversarial settings.
This paper investigates whether early-token confidence signals from LLM decoding can predict reasoning quality in multi-agent debate systems, finding that confidence in the first few generated tokens is the strongest predictor of rubric-based essay scores.
This paper studies the relationship between token-level log-probability distributions, LLM-as-judge rubric scores, and final task accuracy in multi-agent debate systems. It finds a consistent four-phase confidence trajectory and role asymmetry between Constructor and Auditor agents.