Google Research·· 2026-07-16
Towards demystifying the creativity of diffusion models
Towards demystifying the creativity of diffusion models
AI summary
A Google Research ICLR 2026 paper traces diffusion model creativity to score smoothing from regularisation such as weight decay. Smoother learned score functions let denoising interpolate between training points, generating new samples rather than simply memorising them.
Selection record
Threshold 60Official, first-handFirst 31Second 31
Not admittedSum of both 62 < twice the threshold 120
- Source tier
- Official, first-hand; this tier's threshold is 60
- Pre-filter
- passed：探讨扩散模型生成创造力机制
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Google Research · research.google