Imagen de Portada

On the Learning Dynamics of Two-layer Linear Networks with Label Noise SGD

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Zhang, Tongcheng, Zhou, Zhanpeng, Wang, Mingze, Han, Andi, Huang, Wei, Suzuki, Taiji, Yan, Junchi
Formato:	Preprint
Publicado:	2026
Materias:	Machine Learning Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

On the Role of Label Noise in the Feature Learning Process
por: Han, Andi, et al.
Publicado: (2025)

Provably Learning Diffusion Models under the Manifold Hypothesis: Collapse and Refine
por: Huang, Wei, et al.
Publicado: (2026)

On the Emergence of Cross-Task Linearity in the Pretraining-Finetuning Paradigm
por: Zhou, Zhanpeng, et al.
Publicado: (2024)

On the Optimization and Generalization of Two-layer Transformers with Sign Gradient Descent
por: Li, Bingrui, et al.
Publicado: (2024)

The Sharpness Disparity Principle in Transformers for Accelerating Language Model Pre-Training
por: Wang, Jinbo, et al.
Publicado: (2025)

DPRM: A Plug-in Doob h transform-induced Token-Ordering Module for Diffusion Language Models
por: Bu, Dake, et al.
Publicado: (2026)

Provable In-Context Vector Arithmetic via Retrieving Task Concepts
por: Bu, Dake, et al.
Publicado: (2025)

Post-Training as Reweighting: A Stochastic View of Reasoning Trajectories in Language Models
por: Bu, Dake, et al.
Publicado: (2025)

New Evidence of the Two-Phase Learning Dynamics of Neural Networks
por: Zhou, Zhanpeng, et al.
Publicado: (2025)

Learning to Solve Combinatorial Optimization under Positive Linear Constraints via Non-Autoregressive Neural Networks
por: Wang, Runzhong, et al.
Publicado: (2024)

Stochastic Resetting Mitigates Latent Gradient Bias of SGD from Label Noise
por: Bae, Youngkyoung, et al.
Publicado: (2024)

Metastable Dynamics of Chain-of-Thought Reasoning: Provable Benefits of Search, RL and Distillation
por: Kim, Juno, et al.
Publicado: (2025)

How Does Label Noise Gradient Descent Improve Generalization in the Low SNR Regime?
por: Huang, Wei, et al.
Publicado: (2025)

Handling Label Noise via Instance-Level Difficulty Modeling and Dynamic Optimization
por: Zhang, Kuan, et al.
Publicado: (2025)

Stochastic Collapse: How Gradient Noise Attracts SGD Dynamics Towards Simpler Subnetworks
por: Chen, Feng, et al.
Publicado: (2023)

Hide and Seek in Noise Labels: Noise-Robust Collaborative Active Learning with LLM-Powered Assistance
por: Yuan, Bo, et al.
Publicado: (2025)

How Transformers Learn to Plan via Multi-Token Prediction
por: Huang, Jianhao, et al.
Publicado: (2026)

SE-Merging: A Self-Enhanced Approach for Dynamic Model Merging
por: Chen, Zijun, et al.
Publicado: (2025)

SGFormer: Single-Layer Graph Transformers with Approximation-Free Linear Complexity
por: Wu, Qitian, et al.
Publicado: (2024)

Scaling Laws of SignSGD in Linear Regression: When Does It Outperform SGD?
por: Kim, Jihwan, et al.
Publicado: (2026)

Learning under Temporal Label Noise
por: Nagaraj, Sujay, et al.
Publicado: (2024)

DC-SGD: Differentially Private SGD with Dynamic Clipping through Gradient Norm Distribution Estimation
por: Wei, Chengkun, et al.
Publicado: (2025)

FedFixer: Mitigating Heterogeneous Label Noise in Federated Learning
por: Ji, Xinyuan, et al.
Publicado: (2024)

Mean-field Analysis on Two-layer Neural Networks from a Kernel Perspective
por: Takakura, Shokichi, et al.
Publicado: (2024)

Minibatch and Local SGD: Algorithmic Stability and Linear Speedup in Generalization
por: Lei, Yunwen, et al.
Publicado: (2023)

Impact of Label Noise on Learning Complex Features
por: Vashisht, Rahul, et al.
Publicado: (2024)

VocabTailor: Dynamic Vocabulary Selection for Downstream Tasks in Small Language Models
por: Zhang, Hanling, et al.
Publicado: (2025)

A Relative-Budget Theory for Reinforcement Learning with Verifiable Rewards in Large Language Model Reasoning
por: Wachi, Akifumi, et al.
Publicado: (2026)

Robust Deep Hawkes Process under Label Noise of Both Event and Occurrence
por: Tan, Xiaoyu, et al.
Publicado: (2024)

Resurrecting Label Propagation for Graphs with Heterophily and Label Noise
por: Cheng, Yao, et al.
Publicado: (2023)

The Malignant Tail: Spectral Segregation of Label Noise in Over-Parameterized Networks
por: Wang, Zice
Publicado: (2026)

PTCL: Pseudo-Label Temporal Curriculum Learning for Label-Limited Dynamic Graph
por: Zhang, Shengtao, et al.
Publicado: (2025)

Learning Divergence Fields for Shift-Robust Graph Representations
por: Wu, Qitian, et al.
Publicado: (2024)

Addressing Long-Tail Noisy Label Learning Problems: a Two-Stage Solution with Label Refurbishment Considering Label Rarity
por: Wu, Ying-Hsuan, et al.
Publicado: (2024)

Sharpness-Aware Minimization Efficiently Selects Flatter Minima Late in Training
por: Zhou, Zhanpeng, et al.
Publicado: (2024)

NTKMTL: Mitigating Task Imbalance in Multi-Task Learning from Neural Tangent Kernel Perspective
por: Qin, Xiaohan, et al.
Publicado: (2025)

RQP-SGD: Differential Private Machine Learning through Noisy SGD and Randomized Quantization
por: Feng, Ce, et al.
Publicado: (2024)

GSINA: Improving Subgraph Extraction for Graph Invariant Learning via Graph Sinkhorn Attention
por: Yan, Junchi, et al.
Publicado: (2024)

MLLM Is a Strong Reranker: Advancing Multimodal Retrieval-augmented Generation via Knowledge-enhanced Reranking and Noise-injected Training
por: Chen, Zhanpeng, et al.
Publicado: (2024)

Transformers from Diffusion: A Unified Framework for Neural Message Passing
por: Wu, Qitian, et al.
Publicado: (2024)