LLM Digest

Story

arxiv_cs_cl · Oct 26, 2026 · paper

Source brief

From Self-Confidence to Statistical Accuracy: Scaling Human-in-the-Loop Evaluation with Direct and LLM-Assisted Auditing

arxiv.orgOct 26, 2026
original source linked

In brief

For large language models (LLMs) and multimodal large language models (MLLMs), model-reported self-confidence is often miscalibrated, yet it may still be useful for efficient evaluation. At the same time, LLM-as-a-jud...

Feed lens

evaluation

Read the original at arxiv.org →Open in live feed Read that day’s brief

From Self-Confidence to Statistical Accuracy: Scaling Human-in-the-Loop Evaluation with Direct and LLM-Assisted Auditing

Earlier in this thread 4 items

Safety and accuracy follow different scaling laws in clinical large language models

Memory or a Clock? Auditing a Memory Advantage in XLand-MiniGrid with Seeds, Maps and Evaluation Noise Apart

What the Trace Shows and What the Label Says: Auditing Behavioral Detectors in Agent Evaluation

BIRD-RL: Scaling Agentic Reinforcement Learning over Stateful Data-Centric Environments