# @davidad — 2026-05-05

♥34 ↻0 · https://x.com/davidad/status/2051647983758807115

I think steering is a good idea for getting diversity of responses in post-training, where the diverse responses are then ranked for DPO/GRPO by a judge model that is not steered.

tags: author:davidad, kind:tweet, thread-context, year:2026