Cover Image

Enhancing Mixture-of-Experts Specialization via Cluster-Aware Upcycling

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Chu, Sanghyeok, Ahn, Pyunghwan, Song, Gwangmo, Kim, SeungHwan, Lee, Honglak, Han, Bohyung
Format:	Preprint
Published:	2026
Subjects:	Computer Vision and Pattern Recognition
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Exploring the Spectrum of Visio-Linguistic Compositionality and Recognition
by: Oh, Youngtaek, et al.
Published: (2024)

Fine-Grained Captioning of Long Videos through Scene Graph Consolidation
by: Chu, Sanghyeok, et al.
Published: (2025)

Beyond the Ground Truth: Enhanced Supervision for Image Restoration
by: Ryou, Donghun, et al.
Published: (2025)

MoE-GRPO: Optimizing Mixture-of-Experts via Reinforcement Learning in Vision-Language Models
by: Ko, Dohwan, et al.
Published: (2026)

Pri4R: Learning World Dynamics for Vision-Language-Action Models with Privileged 4D Representation
by: Kim, Jisoo, et al.
Published: (2026)

ContextMix: A context-aware data augmentation method for industrial visual inspection systems
by: Kim, Hyungmin, et al.
Published: (2024)

GP-4DGS: Probabilistic 4D Gaussian Splatting from Monocular Video via Variational Gaussian Processes
by: Kim, Mijeong, et al.
Published: (2026)

4D Gaussian Splatting in the Wild with Uncertainty-Aware Regularization
by: Kim, Mijeong, et al.
Published: (2024)

Diffusion-Based Image-to-Image Translation by Noise Correction via Prompt Interpolation
by: Lee, Junsung, et al.
Published: (2024)

CuMo: Scaling Multimodal LLM with Co-Upcycled Mixture-of-Experts
by: Li, Jiachen, et al.
Published: (2024)

Compositional Image Synthesis with Inference-Time Scaling
by: Ji, Minsuk, et al.
Published: (2025)

Map the Flow: Revealing Hidden Pathways of Information in VideoLLMs
by: Kim, Minji, et al.
Published: (2025)

Significantly improving zero-shot X-ray pathology classification via fine-tuning pre-trained image-text encoders
by: Jang, Jongseong, et al.
Published: (2022)

MoE-GS: Mixture of Experts for Dynamic Gaussian Splatting
by: Jin, In-Hwan, et al.
Published: (2025)

Diffusion-Based Conditional Image Editing through Optimized Inference with Guidance
by: Lee, Hyunsoo, et al.
Published: (2024)

Leveraging Temporal Contextualization for Video Action Recognition
by: Kim, Minji, et al.
Published: (2024)

Specializing Foundation Models via Mixture of Low-Rank Experts for Comprehensive Head CT Analysis
by: Yoo, Youngjin, et al.
Published: (2026)

View Selection for 3D Captioning via Diffusion Ranking
by: Luo, Tiange, et al.
Published: (2024)

vid-TLDR: Training Free Token merging for Light-weight Video Transformer
by: Choi, Joonmyung, et al.
Published: (2024)

Multilinear Mixture of Experts: Scalable Expert Specialization through Factorization
by: Oldfield, James, et al.
Published: (2024)

Low-Resolution Editing is All You Need for High-Resolution Editing
by: Lee, Junsung, et al.
Published: (2025)

CLIP-MoE: Towards Building Mixture of Experts for CLIP with Diversified Multiplet Upcycling
by: Zhang, Jihai, et al.
Published: (2024)

Task-customized Masked AutoEncoder via Mixture of Cluster-conditional Experts
by: Liu, Zhili, et al.
Published: (2024)

Domain-Specialized Object Detection via Model-Level Mixtures of Experts
by: Pavlitska, Svetlana, et al.
Published: (2026)

Multi-criteria Token Fusion with One-step-ahead Attention for Efficient Vision Transformers
by: Lee, Sanghyeok, et al.
Published: (2024)

EfficientViM: Efficient Vision Mamba with Hidden State Mixer based State Space Duality
by: Lee, Sanghyeok, et al.
Published: (2024)

Cross-Class Feature Augmentation for Class Incremental Learning
by: Kim, Taehoon, et al.
Published: (2023)

STR-Match: Matching SpatioTemporal Relevance Score for Training-Free Video Editing
by: Lee, Junsung, et al.
Published: (2025)

PhysGaia: A Physics-Aware Benchmark with Multi-Body Interactions for Dynamic Novel View Synthesis
by: Kim, Mijeong, et al.
Published: (2025)

MoLEx: Mixture of Layer Experts for Finetuning with Sparse Upcycling
by: Teo, Rachel S. Y., et al.
Published: (2025)

CLIP-UP: A Simple and Efficient Mixture-of-Experts CLIP Training Recipe with Sparse Upcycling
by: Wang, Xinze, et al.
Published: (2025)

Learning with Noisy Labels: Interconnection of Two Expectation-Maximizations
by: Kim, Heewon, et al.
Published: (2024)

A Training-Free Defense Framework for Robust Learned Image Compression
by: Song, Myungseo, et al.
Published: (2024)

Pixel-aligned RGB-NIR Stereo Imaging and Dataset for Robot Vision
by: Kim, Jinnyeong, et al.
Published: (2024)

ERMoE: Eigen-Reparameterized Mixture-of-Experts for Stable Routing and Interpretable Specialization
by: Cheng, Anzhe, et al.
Published: (2025)

Dynamic Full-body Motion Agent with Object Interaction via Blending Pre-trained Modular Controllers
by: Nam, Sanghyeok, et al.
Published: (2026)

Resolving Task Objective Conflicts in Unified Model via Task-Aware Mixture-of-Experts
by: Zhang, Jiaxing, et al.
Published: (2025)

MedMoE: Modality-Specialized Mixture of Experts for Medical Vision-Language Understanding
by: Chopra, Shivang, et al.
Published: (2025)

Robust Multimodal 3D Object Detection via Modality-Agnostic Decoding and Proximity-based Modality Ensemble
by: Cha, Juhan, et al.
Published: (2024)

Merge and Bound: Direct Manipulations on Weights for Class Incremental Learning
by: Kim, Taehoon, et al.
Published: (2025)