Google DeepMind paper trains models to argue about moral dilemmas; answers become more decisive and stable to option reordering
New paper from Google DeepMind on training models to argue well about moral dilemmas.
The trained model became more decisive and stopped changing its answer when options were reordered. Sycophancy and reframing still moved its judgments.
The setup uses real dilemmas fromXSources2
- @omarsar0XNew paper from Google DeepMind on training models to argue well about moral dilemmas. The trained model became more decisive and stopped changing its answer when options were reordered. Sycophancy and reframing still moved its judgments. The setup uses real dilemmas fromView original ·
arXivResearchConsistency Is Not Validity: Self-Rewarded Moral Argumentation Makes Language Models Decisive but PliableView original ·
Headline by dalao.ai. The quoted post is the record; its author keeps authorship.