Image de couverture de livre

Auto-FlexSwitch: Efficient Dynamic Model Merging via Learnable Task Vector Compression

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Gao, Junqi, Zhang, Dazhi, Guo, Zhichang, Qi, Biqing, Ran, Yi, Zuo, Wangmeng
Format:	Preprint
Publié:	2026
Sujets:	Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

PDAC: Efficient Coreset Selection for Continual Learning via Probability Density Awareness
par: Gao, Junqi, et autres
Publié: (2025)

Less is More: Efficient Model Merging with Binary Task Switch
par: Qi, Biqing, et autres
Publié: (2024)

Perturbation Towards Easy Samples Improves Targeted Adversarial Transferability
par: Gao, Junqi, et autres
Publié: (2024)

Bohdi: Heterogeneous LLM Fusion with Automatic Data Exploration
par: Gao, Junqi, et autres
Publié: (2025)

Evaluating Similitude and Robustness of Deep Image Denoising Models via Adversarial Attack
par: Ning, Jie, et autres
Publié: (2023)

An Efficient Memory Module for Graph Few-Shot Class-Incremental Learning
par: Li, Dong, et autres
Publié: (2024)

Task Vector Quantization for Memory-Efficient Model Merging
par: Kim, Youngeun, et autres
Publié: (2025)

Enhancing Adversarial Transferability via Information Bottleneck Constraints
par: Qi, Biqing, et autres
Publié: (2024)

MergeMoE: Efficient Compression of MoE Models via Expert Output Merging
par: Miao, Ruijie, et autres
Publié: (2025)

Stay Unique, Stay Efficient: Preserving Model Personality in Multi-Task Merging
par: Guo, Kuangpu, et autres
Publié: (2025)

Efficient Visual Transformer by Learnable Token Merging
par: Wang, Yancheng, et autres
Publié: (2024)

MergeMix: Optimizing Mid-Training Data Mixtures via Learnable Model Merging
par: Wang, Jiapeng, et autres
Publié: (2026)

SMR: State Memory Replay for Long Sequence Modeling
par: Qi, Biqing, et autres
Publié: (2024)

DisTaC: Conditioning Task Vectors via Distillation for Robust Model Merging
par: Yoshida, Kotaro, et autres
Publié: (2025)

Exploring Adversarial Robustness of Deep State Space Models
par: Qi, Biqing, et autres
Publié: (2024)

Experimental Analysis of Large-scale Learnable Vector Storage Compression
par: Zhang, Hailin, et autres
Publié: (2023)

FlexQuant: A Flexible and Efficient Dynamic Precision Switching Framework for LLM Quantization
par: Liu, Fangxin, et autres
Publié: (2025)

Task Singular Vectors: Reducing Task Interference in Model Merging
par: Gargiulo, Antonio Andrea, et autres
Publié: (2024)

Localizing Task Information for Improved Model Merging and Compression
par: Wang, Ke, et autres
Publié: (2024)

Online DPO: Online Direct Preference Optimization with Fast-Slow Chasing
par: Qi, Biqing, et autres
Publié: (2024)

LPT++: Efficient Training on Mixture of Long-tailed Experts
par: Dong, Bowen, et autres
Publié: (2024)

Fast and Slow Gradient Approximation for Binary Neural Network Optimization
par: Chen, Xinquan, et autres
Publié: (2024)

Whoever Started the Interference Should End It: Guiding Data-Free Model Merging via Task Vectors
par: Cheng, Runxi, et autres
Publié: (2025)

Deciphering the Chaos: Enhancing Jailbreak Attacks via Adversarial Prompt Translation
par: Li, Qizhang, et autres
Publié: (2024)

WIST: Web-Grounded Iterative Self-Play Tree for Domain-Targeted Reasoning Improvement
par: Li, Fangyuan, et autres
Publié: (2026)

Sub-MoE: Efficient Mixture-of-Expert LLMs Compression via Subspace Expert Merging
par: Li, Lujun, et autres
Publié: (2025)

Vectorized Attention with Learnable Encoding for Quantum Transformer
par: Guo, Ziqing, et autres
Publié: (2025)

Interactive Continual Learning: Fast and Slow Thinking
par: Qi, Biqing, et autres
Publié: (2024)

FlexRound: Learnable Rounding based on Element-wise Division for Post-Training Quantization
par: Lee, Jung Hyun, et autres
Publié: (2023)

AdaMerging: Adaptive Model Merging for Multi-Task Learning
par: Yang, Enneng, et autres
Publié: (2023)

Localize-and-Stitch: Efficient Model Merging via Sparse Task Arithmetic
par: He, Yifei, et autres
Publié: (2024)

AutoCompress: Critical Layer Isolation for Efficient Transformer Compression
par: Thorat, Archit
Publié: (2026)

Superpose Task-specific Features for Model Merging
par: Qiu, Haiquan, et autres
Publié: (2025)

Improved Generation of Adversarial Examples Against Safety-aligned LLMs
par: Li, Qizhang, et autres
Publié: (2024)

Multi-Task Model Merging via Adaptive Weight Disentanglement
par: Xiong, Feng, et autres
Publié: (2024)

Improving Transferability of Adversarial Examples via Bayesian Attacks
par: Li, Qizhang, et autres
Publié: (2023)

Efficient and Effective Weight-Ensembling Mixture of Experts for Multi-Task Model Merging
par: Shen, Li, et autres
Publié: (2024)

Tiled Bit Networks: Sub-Bit Neural Network Compression Through Reuse of Learnable Binary Vectors
par: Gorbett, Matt, et autres
Publié: (2024)

Task Vector Bases: A Unified and Scalable Framework for Compressed Task Arithmetic
par: Zeng, Siqi, et autres
Publié: (2025)

SpecMol: A Spectroscopy-Grounded Foundation Model for Multi-Task Molecular Learning
par: Shen, Shuaike, et autres
Publié: (2025)