Verified70% confidenceFactExact time
Knowledge distillation, proposed by Hinton et al., has student models inherit implicit knowledge by learning from the soft probability distributions output by teacher models.
4
Sources
70%
Confidence
Long-term
Relevance
7/16/2026
First Seen
Sources
Related Claims
Verified知识蒸馏由Hinton等人在2015年提出，核心思想是让小型学生模型模仿大型教师模型的输出分布83% similarUnverified知识蒸馏由Hinton等人2015年提出，通过温度参数软化概率分布，使低概率类别的暗知识更显著供学生模型学习83% similarVerifiedThe core idea of knowledge distillation is to have the student model imitate the soft labels (probability distributions across categories) output by the teacher model, with soft labels being referred to by Hinton as "dark knowledge".83% similarUnverifiedKnowledge distillation was first proposed by Hinton, Vinyals, and Dean at Google in 2015, with the core idea of having a small student model learn from the soft labels output by a large teacher model.75% similarUnverified知识蒸馏最早由Hinton等人在2015年提出，DiffusionGemma创造性地将其用于步数压缩，教师模型用较多步数（如48步）生成，学生模型用更少步数（如12步）达到相近质量72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/533431API
curl https://kongchang.com/api/v1/knowledge/claims/533431MCP
get_claim(id=533431)