Unverified50% confidenceFactExact time
2020年Jonathan Ho等人发表的DDPM论文奠定了现代扩散模型的基础框架，核心思路是逐步加入噪声再训练神经网络学习去噪的逆过程
1
Sources
50%
Confidence
Long-term
Relevance
7/13/2026
First Seen
Sources
朴素艺术与AI时代：Gina画廊如何定义真实创作
hackernewshackernews7/12/2026
Related Claims
UnverifiedGoogle's Noisy Student Training demonstrated around 2020 the effectiveness of using a teacher model to generate pseudo-labels for training a student model.58% similarUnverifiedKnowly's active learning feature draws on the learning science principle of Spaced Repetition and combines it with the contextual understanding capabilities of LLMs.55% similarUnverified斯坦福的ACT和CMU的Diffusion Policy验证了端到端感知-动作学习在灵巧操作任务中的有效性，Diffusion Policy将DDPM引入动作生成54% similarUnverifiedC51（Categorical DQN）由Bellemare、Dabney和Munos于2017年提出，是分布式强化学习的开山之作54% similarUnverifiedAndrew Ng proposed the strategy "build an end-to-end system first, then iterate to improve" in his book Machine Learning Yearning.53% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/502850API
curl https://kongchang.com/api/v1/knowledge/claims/502850MCP
get_claim(id=502850)