Unverified50% confidenceOpinionExact time
Noam Brown pointed out that the relationship between different scaling axes is multiplicative rather than additive, and that there is a multiplicative effect between reinforcement learning and pretraining
1
Sources
50%
Confidence
Long-term
Relevance
10/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedSmith and Le (2018) showed that the ratio of learning rate to batch size η/B controls the strength of implicit regularization.66% similarUnverified融合符号搜索与梯度学习的推荐学习路径为先打好优化理论与逻辑类型论双侧基础，再以范畴论为桥梁，最后用前沿论文驱动学习64% similarUnverified元强化学习是'学会学习'研究范式在强化学习领域的具体体现63% similarUnverifiedThe paper identifies reinforcement learning as a promising direction for automated alignment testing methods that scale with compute while efficiently utilizing compute.62% similarUnverified双环学习概念由管理学家克里斯·阿吉里斯提出，强调组织不仅从结果中学习还要反思修正底层假设62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/972618API
curl https://kongchang.com/api/v1/knowledge/claims/972618MCP
get_claim(id=972618)