Copertina

A Unified Generalization Framework for Model Merging: Trade-offs, Non-Linearity, and Scaling Laws

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Li, Qinglun, Tang, Anke, Zhang, Miao, Wang, Mengzhu, Yin, Quanjun, Shen, Li
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

OledFL: Unleashing the Potential of Decentralized Federated Learning via Opposite Lookahead Enhancement
di: Li, Qinglun, et al.
Pubblicazione: (2024)

Boosting the Performance of Decentralized Federated Learning via Catalyst Acceleration
di: Li, Qinglun, et al.
Pubblicazione: (2024)

Unveiling the Power of Multiple Gossip Steps: A Stability-Based Generalization Analysis in Decentralized Training
di: Li, Qinglun, et al.
Pubblicazione: (2025)

PACE: Parameter Change for Unsupervised Environment Design
di: Yuan, Fang, et al.
Pubblicazione: (2026)

Merging Multi-Task Models via Weight-Ensembling Mixture of Experts
di: Tang, Anke, et al.
Pubblicazione: (2024)

Merging Models on the Fly Without Retraining: A Sequential Approach to Scalable Continual Model Merging
di: Tang, Anke, et al.
Pubblicazione: (2025)

Modeling Multi-Task Model Merging as Adaptive Projective Gradient Descent
di: Wei, Yongxian, et al.
Pubblicazione: (2025)

Mitigating the Backdoor Effect for Multi-Task Model Merging via Safety-Aware Subspace
di: Yang, Jinluan, et al.
Pubblicazione: (2024)

Merge and Guide: Unifying Model Merging and Guided Decoding for Controllable Multi-Objective Generation
di: Xie, Guofu, et al.
Pubblicazione: (2025)

Trade-offs in Ensembling, Merging and Routing Among Parameter-Efficient Experts
di: Lotfi, Sanae, et al.
Pubblicazione: (2026)

HydraOpt: Navigating the Efficiency-Performance Trade-off of Adapter Merging
di: Ceritli, Taha, et al.
Pubblicazione: (2025)

Does Scaling Law Apply in Time Series Forecasting?
di: Li, Zeyan, et al.
Pubblicazione: (2025)

Scaling Laws for Precision in High-Dimensional Linear Regression
di: Zhang, Dechen, et al.
Pubblicazione: (2026)

Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Merging
di: Yang, Jinluan, et al.
Pubblicazione: (2025)

Breaking the Likelihood-Quality Trade-off in Diffusion Models by Merging Pretrained Experts
di: Esfandiari, Yasin, et al.
Pubblicazione: (2025)

Efficient and Effective Weight-Ensembling Mixture of Experts for Multi-Task Model Merging
di: Shen, Li, et al.
Pubblicazione: (2024)

Understanding Generalization of Federated Learning: the Trade-off between Model Stability and Optimization
di: Zeng, Dun, et al.
Pubblicazione: (2024)

Parameter Efficient Multi-task Model Fusion with Partial Linearization
di: Tang, Anke, et al.
Pubblicazione: (2023)

Efficient Linear Attention for Multivariate Time Series Modeling via Entropy Equality
di: Zhang, Mingtao, et al.
Pubblicazione: (2025)

Merging Smarter, Generalizing Better: Enhancing Model Merging on OOD Data
di: Zhang, Bingjie, et al.
Pubblicazione: (2025)

Improving the Trade-off Between Watermark Strength and Speculative Sampling Efficiency for Language Models
di: He, Weiqing, et al.
Pubblicazione: (2026)

Scaling Laws and In-Context Learning: A Unified Theoretical Framework
di: Mehta, Sushant, et al.
Pubblicazione: (2025)

A Unified Blockwise Measurement Design for Learning Quantum Channels and Lindbladians via Low-Rank Matrix Sensing
di: Lang, Quanjun, et al.
Pubblicazione: (2025)

Learning Memory Kernels in Generalized Langevin Equations
di: Lang, Quanjun, et al.
Pubblicazione: (2024)

A Unified Framework for Knowledge Transfer in Bidirectional Model Scaling
di: Shen, Jianlu, et al.
Pubblicazione: (2026)

The Risk of Federated Learning to Skew Fine-Tuning Features and Underperform Out-of-Distribution Robustness
di: Du, Mengyao, et al.
Pubblicazione: (2024)

A Unified and General Framework for Continual Learning
di: Wang, Zhenyi, et al.
Pubblicazione: (2024)

Provable Sparse Inversion and Token Relabel Enhanced One-shot Federated Learning with ViTs
di: Shen, Li, et al.
Pubblicazione: (2026)

Navigating the Alignment-Calibration Trade-off: A Pareto-Superior Frontier via Model Merging
di: Hu, Tiancheng, et al.
Pubblicazione: (2025)

Beyond the False Trade-off: Adaptive EWC for Stealthy and Generalizable T2I Backdoors
di: Bowen, Lu, et al.
Pubblicazione: (2026)

CoCo: A Coupled Contrastive Framework for Unsupervised Domain Adaptive Graph Classification
di: Yin, Nan, et al.
Pubblicazione: (2023)

Differentially Private Model Merging
di: Yin, Qichuan, et al.
Pubblicazione: (2026)

Unified Neural Network Scaling Laws and Scale-time Equivalence
di: Boopathy, Akhilan, et al.
Pubblicazione: (2024)

GRE^2-MDCL: Graph Representation Embedding Enhanced via Multidimensional Contrastive Learning
di: Fan, Kaizhe, et al.
Pubblicazione: (2024)

A Unified Learn-to-Distort-Data Framework for Privacy-Utility Trade-off in Trustworthy Federated Learning
di: Zhang, Xiaojin, et al.
Pubblicazione: (2024)

USBD: Universal Structural Basis Distillation for Source-Free Graph Domain Adaptation
di: Wang, Yingxu, et al.
Pubblicazione: (2026)

Unifying Learning Dynamics and Generalization in Transformers Scaling Law
di: Yang, Chiwun
Pubblicazione: (2025)

AdaMerging: Adaptive Model Merging for Multi-Task Learning
di: Yang, Enneng, et al.
Pubblicazione: (2023)

FusionBench: A Unified Library and Comprehensive Benchmark for Deep Model Fusion
di: Tang, Anke, et al.
Pubblicazione: (2024)

Unified Scaling Laws for Compressed Representations
di: Panferov, Andrei, et al.
Pubblicazione: (2025)