InferenceMAX Evaluates Wh Per Query For LLMs

The item asks whether the InferenceMAX benchmark dataset can reasonably be used to derive watt-hour (Wh) per-query energy consumption figures for large language models (LLMs). It frames a feasibility question about using that dataset for per-query energy estimates.