Unverified50% confidenceFactExact time
规模缩放定律研究表明，增大模型参数量能提升常见知识准确性，但对长尾事实的改善边际效益递减
1
Sources
50%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
Related Claims
Unverified丰富的上下文信息能收窄模型输出概率分布并减少幻觉，这一现象被称为「知识锚定效应」77% similarUnverifiedOpenAI、DeepMind等机构的研究表明模型性能随参数量、数据量和计算量的增加呈幂律提升（Scaling Law）74% similarUnverifiedReasoning capabilities can be compressed through knowledge distillation and reinforcement learning fine-tuning, but the storage of factual knowledge depends on parameter scale, and compression leads to reduced knowledge coverage and higher hallucination rates.74% similarUnverified结构化、上下文丰富的提示不仅提升模型输出质量，也能显著降低安全分类器的误触发率71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/489863API
curl https://kongchang.com/api/v1/knowledge/claims/489863MCP
get_claim(id=489863)