Unverified90% confidenceFactExact time
Finding the function mapping in supervised learning is essentially an optimization problem that adjusts model parameters by minimizing the loss function (difference between predicted and true values).
1
Sources
90%
Confidence
Long-term
Relevance
8/2/2026
First Seen
Sources
Related Entities
Related Claims
Unverified分类模型的推理是非自回归的，仅需一次前向传播即可完成所有位置预测，计算效率显著高于生成模型67% similarUnverified监督微调训练模型模仿人类标注的正确答案，而强化学习微调让模型通过试错学习最优策略66% similarPartially VerifiedThe goal of supervised learning is to find a function f that, for a given input x, can predict the corresponding output y.66% similarUnverified预训练的目标函数通常是因果语言建模中的交叉熵损失，即给定前文所有token最大化正确预测下一个token的概率65% similarVerified微调是在通用预训练模型基础上用特定领域数据进行二次训练，使模型能将领域知识内化为参数，推理时无需额外检索步骤，延迟更低65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/679933API
curl https://kongchang.com/api/v1/knowledge/claims/679933MCP
get_claim(id=679933)