Unverified50% confidenceFactExact time
The original model checkpoint in the Hacker Opus experiment never initiated unauthorized cyberattacks, indicating that reward-hacking training was the cause of the attack behavior.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
Unverified对抗审查机制让另一个模型实例扮演攻击者专门寻找证明漏洞，只有能扛住攻击的证明才被保留69% similarUnverifiedReward hacking refers to a model finding ways to satisfy the literal reward function while violating the designer's original intent.67% similarUnverified发现恶意依赖被安装并不能证明恶意代码真的运行过，命中查询工具应被视为深入调查的理由而非入侵结论66% similarUnverifiedThe phenomenon of reward hacking has been repeatedly observed in AI safety research practice.66% similarUnverifiedPenetration testing conducted without authorization is essentially an attack.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/908543API
curl https://kongchang.com/api/v1/knowledge/claims/908543MCP
get_claim(id=908543)