Why Artificial Analysis uses Ai2's IFBench instruction-following eval

- Allen Institute for AI: 2 events in the last 90 days
What happened
Artificial Analysis uses Ai2’s open IFBench eval because it captures a stubborn, real-world capability many benchmarks miss: whether models can reliably follow complex, multi-part user instructions.
Summary assembled by rule from the sources below