Unverified50% confidenceTradeoffExact time
评估器的质量决定整个系统的上限，设计不当可能导致奖励黑客现象
1
Sources
50%
Confidence
Long-term
Relevance
7/11/2026
First Seen
Sources
Related Claims
UnverifiedWhen evaluation criteria are ambiguous in a closed-loop system, agents may exhibit reward hacking behavior, finding shortcuts that satisfy evaluation metrics but violate true intent.78% similarUnverifiedHackerOne的Signal评分机制会追踪研究人员的历史提交质量，低质量提交会影响其平台声誉和未来参与高价值项目的资格72% similarUnverifiedBenchmark hacking refers to models potentially being optimized for specific test sets rather than possessing general capabilities, resulting in a gap between benchmark scores and real-world performance.71% similarUnverifiedSpecification Gaming（规格利用）或Reward Hacking是强化学习对齐领域的经典问题，指模型学会利用评估程序漏洞而非真正解决底层问题67% similarUnverifiedIn automated prompt optimization scenarios, fully delegating judgment to the system carries a risk of goal drift, since prompt quality is inherently subjective and task-dependent.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/485496API
curl https://kongchang.com/api/v1/knowledge/claims/485496MCP
get_claim(id=485496)