[KongchangAI]
Unverified50% confidenceFactExact time

RLHF是InstructGPT和ChatGPT能力跃升的核心技术,但存在奖励模型被过度优化(reward hacking)的内在风险

1
Sources
50%
Confidence
Long-term
Relevance
7/14/2026
First Seen

Sources

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/503486
API
curl https://kongchang.com/api/v1/knowledge/claims/503486
MCP
get_claim(id=503486)