Fetching the paper…

STEER-BENCH: A Benchmark for Evaluating the Steerability of Large Language Models · Around