Unverified60% confidenceFactExact time
Reward shaping techniques were theoretically analyzed by Andrew Ng and others in a 1999 paper, which proved how reward shaping can be safely performed without altering the optimal policy.
2
Sources
60%
Confidence
Long-term
Relevance
8/9/2026
First Seen
Sources
Related Claims
Unverified1999年Ng、Harada与Russell在ICML证明，只有基于势函数的奖励塑形（F(s,s')=γΦ(s')-Φ(s)）才能保证不改变最优策略62% similarUnverified基于势函数的奖励塑形（Ng 等人，1999）经理论证明不会改变最优策略，其形式为 F = γΦ(s') - Φ(s)62% similarUnverifiedThe study characterizes the asymptotic behavior of the resulting exploration criterion at the theoretical level and derives the performance of the actual physical system under the optimized policy.60% similarUnverified斯图尔特·罗素在2019年出版的《与人共生》（Human Compatible）中系统论证了不确定性回报函数的重要性60% similarUnverifiedOpenAI claims its internal model has solved the Millennium Prize problem related to the Navier-Stokes equations, providing both an analytical proof and a Lean formal proof.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/718198API
curl https://kongchang.com/api/v1/knowledge/claims/718198MCP
get_claim(id=718198)