LLM Digest

Story

arxiv_llm_reliability · Sep 28, 2026 · paper

Source brief

Does Model Uncertainty Track Human Ambiguity? Evidence from Multi-Annotator Vision Benchmarks

arxiv.orgSep 28, 2026
original source linked

In brief

Human-model alignment is critical for trustworthy AI-assisted decision-making systems. Yet, most work evaluates model predictions against single ground-truth labels, overlooking that humans themselves often disagree o...

Feed lens

eval

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

Does Model Uncertainty Track Human Ambiguity? Evidence from Multi-Annotator Vision Benchmarks

Earlier in this thread 4 items

The Missing Complement: State-Conditioned Minimal Sufficient Evidence for Coding Agents

REVA: Reusable Evidence View Aggregation for Context-Efficient RAG Serving

UTILMEM: Benchmarking Evidence Utilization in Long-Term Conversational Memory

Lazy Grounding: Attacking Search Agents with Factual Evidence