Benchmark
AdvancedIF
Evaluate complex, multi-turn, and system-level instruction following with expert-written rubrics.
Published leaderboard
Explore scores and evaluation settings on the publisher’s leaderboard.
Open leaderboardCoverage
More than 1,600 instruction-following tasks assessed with expert-curated rubrics.
Methodology
Expert-written rubrics measure compliance with complex instructions. Mercor evaluates the same benchmark through its own harness; its results are evaluator-specific. RIFL is the paper’s separate training method.