Unverified50% confidenceFactExact time
Designing reward functions that fully capture human intent is extremely difficult, a challenge the academic community refers to as the Reward Design Problem.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedSparse rewards are one of the most challenging problems in reinforcement learning, where an agent only receives a reward upon completing a specific objective.72% similarUnverified编程任务的正确性有客观执行结果作为奖励信号，而创意类任务的奖励函数需人为构造，导致AI在创意类工作上表现不稳定70% similarVerified奖励函数设计不当会导致奖励欺骗（Reward Hacking）现象，即智能体找到在奖励函数上得分高但实际无意义甚至有害的行为路径68% similarUnverified在设计约束下有限资源能催生创新的现象被称为'约束悖论'（Constraints Paradox）67% similarUnverifiedStructural errors can be resolved with engineering constraints, but taste and judgment at the semantic level currently remain difficult to correct through simple rules and require human intervention.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/893061API
curl https://kongchang.com/api/v1/knowledge/claims/893061MCP
get_claim(id=893061)