confidence deep-learning evidence neural-network neurips-2020 pytorch tensorflow uncertainty

adaptive adoptr confidence designs estimation intervals optimal parameter point two-stage

confidence factual factuality llm llms robustness

confidence effect interval size statistics