Mikhail Belkin
~Mikhail_Belkin1
12
论文总数
6.0
年均投稿
平均评分
接收情况7/12
会议分布
ICLR
7
NeurIPS
3
ICML
2
发表论文 (12 篇)
20256 篇
4
Seeds of Structure: Patch PCA Reveals Universal Compositional Cues in Diffusion Models
NeurIPS 2025Poster
5
Fast Training of Large Kernel Models with Delayed Projections
NeurIPS 2025Spotlight
4
Context-Scaling versus Task-Scaling in In-Context Learning
ICLR 2025Rejected
3
Emergence in non-neural models: grokking modular arithmetic via average gradient outer product
ICLR 2025Rejected
4
Emergence in non-neural models: grokking modular arithmetic via average gradient outer product
ICML 2025Oral
4
Task Generalization with Autoregressive Compositional Structure: Can Learning from $D$ Tasks Generalize to $D^T$ Tasks?
ICML 2025Poster
20246 篇
4
Mechanism of clean-priority learning in early stopped neural networks of infinite width
ICLR 2024Rejected
4
SGD batch saturation for training wide neural networks
ICLR 2024Rejected
5
Quadratic models for understanding catapult dynamics of neural networks
ICLR 2024Poster
3
More is Better: when Infinite Overparameterization is Optimal and Overfitting is Obligatory
ICLR 2024Poster
4
Average gradient outer product as a mechanism for deep neural collapse
NeurIPS 2024Poster
4
Catapults in SGD: spikes in the training loss and their impact on generalization through feature learning
ICLR 2024Rejected