Answers are grounded in this paper and cite their sources.
Preprint evaluates latency and health labels in LLM endpoints, highlighting measurement limitations.
This preprint presents a black-box measurement study of publicly accessible LLM inference endpoints. Using approximately 64,000 production telemetry probes collected over nine days, the study evaluates latency behavior, right-censored observations, threshold-derived health labels, and measurement limitations in operational monitoring systems.
No takes yet. Share an insight, caveat, or question.
Aathil Felix Clarisa (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: