gemma.cpp/build/_deps/benchmark-src/docs/random_interleaving.md

Random Interleaving

Random Interleaving is a technique to lower run-to-run variance. It randomly interleaves repetitions of a microbenchmark with repetitions from other microbenchmarks in the same benchmark test. Data shows it is able to lower run-to-run variance by 40% on average.

To use, you mainly need to set --benchmark_enable_random_interleaving=true, and optionally specify non-zero repetition count --benchmark_repetitions=9 and optionally decrease the per-repetition time --benchmark_min_time=0.1.