Copertina

A Statistical Theory of Gated Attention through the Lens of Hierarchical Mixture of Experts

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Nguyen, Viet, Pham, Tuan Minh, Cao, Thinh, Dinh, Tan, Nguyen, Huy, Ho, Nhat, Rinaldo, Alessandro
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Rethinking Multinomial Logistic Mixture of Experts with Sigmoid Gating Function
di: Pham, Tuan Minh, et al.
Pubblicazione: (2026)

Convergence Rates for Softmax Gating Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2025)

On Least Square Estimation in Softmax Gating Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2024)

Sigmoid Gating is More Sample Efficient than Softmax Gating in Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2024)

On Bayesian Softmax-Gated Mixture-of-Experts Models
di: Bariletto, Nicola, et al.
Pubblicazione: (2026)

Quadratic Gating Mixture of Experts: Statistical Insights into Self-Attention
di: Akbarian, Pedram, et al.
Pubblicazione: (2024)

On DeepSeekMoE: Statistical Benefits of Shared Experts and Normalized Sigmoid Gating
di: Nguyen, Huy, et al.
Pubblicazione: (2025)

A General Theory for Softmax Gating Multinomial Logistic Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2023)

Sigmoid Self-Attention has Lower Sample Complexity than Softmax Self-Attention: A Mixture-of-Experts Perspective
di: Yan, Fanqi, et al.
Pubblicazione: (2025)

Statistical Perspective of Top-K Sparse Softmax Gating Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2023)

On Expert Estimation in Hierarchical Mixture of Experts: Beyond Softmax Gating Functions
di: Nguyen, Huy, et al.
Pubblicazione: (2024)

Statistical Advantages of Perturbing Cosine Router in Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2024)

On Minimax Estimation of Parameters in Softmax-Contaminated Mixture of Experts
di: Yan, Fanqi, et al.
Pubblicazione: (2025)

On Parameter Estimation in Deviated Gaussian Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2024)

Mixture of Experts Meets Prompt-Based Continual Learning
di: Le, Minh, et al.
Pubblicazione: (2024)

RepLoRA: Reparameterizing Low-Rank Adaptation via the Perspective of Mixture of Experts
di: Truong, Tuan, et al.
Pubblicazione: (2025)

Is Temperature Sample Efficient for Softmax Gaussian Mixture of Experts?
di: Nguyen, Huy, et al.
Pubblicazione: (2024)

Towards Convergence Rates for Parameter Estimation in Gaussian-gated Mixture of Experts
di: Nguyen, Huy, et al.
Pubblicazione: (2023)

Hypernetwork-Driven Low-Rank Adaptation Across Attention Heads
di: Diep, Nghiem T., et al.
Pubblicazione: (2025)

Improving Minimax Estimation Rates for Contaminated Mixture of Multinomial Logistic Experts via Expert Heterogeneity
di: Yan, Fanqi, et al.
Pubblicazione: (2026)

One-Prompt Strikes Back: Sparse Mixture of Experts for Prompt-based Continual Learning
di: Le, Minh, et al.
Pubblicazione: (2025)

Modeling Expert Interactions in Sparse Mixture of Experts via Graph Structures
di: Nguyen-Nhat, Minh-Khoi, et al.
Pubblicazione: (2025)

On Zero-Initialized Attention: Optimal Prompt and Gating Factor Estimation
di: Diep, Nghiem T., et al.
Pubblicazione: (2025)

Understanding Expert Structures on Minimax Parameter Estimation in Contaminated Mixture of Experts
di: Yan, Fanqi, et al.
Pubblicazione: (2024)

FuseMoE: Mixture-of-Experts Transformers for Fleximodal Fusion
di: Han, Xing, et al.
Pubblicazione: (2024)

Revisit Visual Prompt Tuning: The Expressiveness of Prompt Experts
di: Le, Minh, et al.
Pubblicazione: (2025)

Revisiting Prefix-tuning: Statistical Benefits of Reparameterization among Prompts
di: Le, Minh, et al.
Pubblicazione: (2024)

CompeteSMoE -- Effective Training of Sparse Mixture of Experts via Competition
di: Pham, Quang, et al.
Pubblicazione: (2024)

DoRAN: Stabilizing Weight-Decomposed Low-Rank Adaptation via Noise Injection and Auxiliary Networks
di: Diep, Nghiem T., et al.
Pubblicazione: (2025)

Dendrograms of Mixing Measures for Softmax-Gated Gaussian Mixture of Experts: Consistency without Model Sweeps
di: Hai, Do Tien, et al.
Pubblicazione: (2025)

Lightspeed Geometric Dataset Distance via Sliced Optimal Transport
di: Nguyen, Khai, et al.
Pubblicazione: (2025)

Model Selection for Gaussian-gated Gaussian Mixture of Experts Using Dendrograms of Mixing Measures
di: Thai, Tuan, et al.
Pubblicazione: (2025)

Leveraging Hierarchical Taxonomies in Prompt-based Continual Learning
di: Tran, Quyen, et al.
Pubblicazione: (2024)

Expert Merging in Sparse Mixture of Experts with Nash Bargaining
di: Nguyen, Dung V., et al.
Pubblicazione: (2025)

Improving Generalization in Heterogeneous Federated Continual Learning via Spatio-Temporal Gradient Matching with Prototypical Coreset
di: Nguyen, Minh-Duong, et al.
Pubblicazione: (2025)

CompeteSMoE -- Statistically Guaranteed Mixture of Experts Training via Competition
di: Nguyen, Nam V., et al.
Pubblicazione: (2025)

Fast Model Selection and Stable Optimization for Softmax-Gated Multinomial-Logistic Mixture of Experts Models
di: Tran, TrungKhang, et al.
Pubblicazione: (2026)

On Linear Mode Connectivity of Mixture-of-Experts Architectures
di: Tran, Viet-Hoang, et al.
Pubblicazione: (2025)

Beyond Vanilla Variational Autoencoders: Detecting Posterior Collapse in Conditional and Hierarchical Variational Autoencoders
di: Dang, Hien, et al.
Pubblicazione: (2023)

SAGE: Shape-Adapting Gated Experts for Adaptive Histopathology Image Segmentation
di: Thai, Gia Huy, et al.
Pubblicazione: (2025)