Unverified50% confidenceFactExact time
古德哈特定律指出当一个度量指标成为优化目标时，它就不再是好的度量指标，在语言模型中体现为奖励黑客现象
1
Sources
50%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
Related Claims
Verified古德哈特定律描述：当一个度量指标变成目标时，它就不再是好的度量指标了80% similarUnverifiedPoorly designed evaluators can lead to reward hacking, where the model finds shortcuts that score high on the metric but are actually useless in practice.77% similarUnverifiedGoodhart's Law states that when a quality metric becomes an optimization target, it is no longer a good measure of quality, as the model may learn to game the score rather than genuinely improve.74% similarUnverified模型优化测量指标本身而非其代表的真实能力的现象在AI安全领域被称为规格博弈（Specification Gaming）或奖励黑客（Reward Hacking）72% similarUnverifiedIn reinforcement learning literature, the phenomenon where a model maximizes contributions to short-term objectives while ignoring implicit constraints is referred to as Reward Hacking or Specification Gaming.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/594853API
curl https://kongchang.com/api/v1/knowledge/claims/594853MCP
get_claim(id=594853)