"feature learning" Papers

16 papers found

From Linear to Nonlinear: Provable Weak-to-Strong Generalization through Feature Learning

Junsoo Oh, Jerry Song, Chulhee Yun

NeurIPS 2025posterarXiv:2510.24812
2
citations

How Data Mixing Shapes In-Context Learning: Asymptotic Equivalence for Transformers with MLPs

Samet Demir, Zafer Dogan

NeurIPS 2025posterarXiv:2510.25753

Learning Hierarchical Polynomials of Multiple Nonlinear Features

Hengyu Fu, Zihao Wang, Eshaan Nichani et al.

ICLR 2025posterarXiv:2411.17201
3
citations

On the Feature Learning in Diffusion Models

Andi Han, Wei Huang, Yuan Cao et al.

ICLR 2025posterarXiv:2412.01021
13
citations

Revisiting Residual Connections: Orthogonal Updates for Stable and Efficient Deep Networks

Giyeong Oh, Woohyun Cho, Siyeol Kim et al.

NeurIPS 2025posterarXiv:2505.11881

The Computational Advantage of Depth in Learning High-Dimensional Hierarchical Targets

Yatin Dandi, Luca Pesce, Lenka Zdeborová et al.

NeurIPS 2025spotlight

Asymptotics of feature learning in two-layer networks after one gradient-step

Hugo Cui, Luca Pesce, Yatin Dandi et al.

ICML 2024spotlight

A Theory of Non-Linear Feature Learning with One Gradient Step in Two-Layer Neural Networks

Behrad Moniri, Donghwan Lee, Hamed Hassani et al.

ICML 2024poster

Catapults in SGD: spikes in the training loss and their impact on generalization through feature learning

Libin Zhu, Chaoyue Liu, Adityanarayanan Radhakrishnan et al.

ICML 2024posterarXiv:2306.04815

Diffusion Models Demand Contrastive Guidance for Adversarial Purification to Advance

Mingyuan Bai, Wei Huang, Li Tenghui et al.

ICML 2024poster

LoRA+: Efficient Low Rank Adaptation of Large Models

Soufiane Hayou, Nikhil Ghosh, Bin Yu

ICML 2024posterarXiv:2402.12354

Mean-field Analysis on Two-layer Neural Networks from a Kernel Perspective

Shokichi Takakura, Taiji Suzuki

ICML 2024poster

Mean Field Langevin Actor-Critic: Faster Convergence and Global Optimality beyond Lazy Learning

Kakei Yamamoto, Kazusato Oko, Zhuoran Yang et al.

ICML 2024oral

Provable Benefits of Local Steps in Heterogeneous Federated Learning for Neural Networks: A Feature Learning Perspective

Yajie Bao, Michael Crawshaw, Mingrui Liu

ICML 2024poster

Provable Multi-Task Representation Learning by Two-Layer ReLU Neural Networks

Liam Collins, Hamed Hassani, Mahdi Soltanolkotabi et al.

ICML 2024poster

Provably Neural Active Learning Succeeds via Prioritizing Perplexing Samples

Dake Bu, Wei Huang, Taiji Suzuki et al.

ICML 2024poster