A Definition of Catastrophic Performance

概要

A central challenge in AI is specifying objectives that align with human preferences. While existing theoretical results elucidate how misspecification impacts performance, they typically do not distinguish between merely suboptimal and catastrophic outcomes. To enable rigorous analysis of catastrophic risk from AI misalignment, we introduce a formal definition of catastrophic performance. Our definition is based on two reference values. The first, baseline value, captures the value attained if no agent is deployed. The second, orthogonal value, captures the value of a trajectory that is unrelated to human preferences, in a sense we describe. Orthogonal value is typically far worse than baseline value. We interpret these reference values through examples in which past human activity has altered the state of the environment to human benefit. Starting at the resulting state, an AI agent is deployed with the goal of accelerating further progress. Catastrophic performance represents this agent's undoing of benefits afforded by human activity.

参考リンク

• Paper ↗

リンク先のライセンス・再現性・データ条件をご確認ください。自動分類や実装の目安は保証ではありません。

実装を始めるプロンプト

参考リンクを確認したうえで、次の内容をコピーして利用できます。

# 論文・記事
A Definition of Catastrophic Performance

# 概要
A central challenge in AI is specifying objectives that align with human preferences. While existing theoretical results elucidate how misspecification impacts performance, they typically do not…

# 原典
https://arxiv.org/abs/2611.06219

上記の手法を実装してください。実行前に依存関係とライセンスを確認してください。