Imagen de Portada

Dataless Weight Disentanglement in Task Arithmetic via Kronecker-Factored Approximate Curvature

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Porrello, Angelo, Buzzega, Pietro, Dangel, Felix, Sommariva, Thomas, Salami, Riccardo, Bonicelli, Lorenzo, Calderara, Simone
Formato:	Preprint
Publicado:	2026
Materias:	Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Rethinking Layer-wise Model Merging through Chain of Merges
por: Buzzega, Pietro, et al.
Publicado: (2025)

Distilling Linearized Behavior into Non-Linear Fine-Tuning for Effective Task Arithmetic
por: Sommariva, Thomas, et al.
Publicado: (2026)

A Second-Order Perspective on Model Compositionality and Incremental Learning
por: Porrello, Angelo, et al.
Publicado: (2024)

Modular Embedding Recomposition for Incremental Learning
por: Panariello, Aniello, et al.
Publicado: (2025)

CLIP with Generative Latent Replay: a Strong Baseline for Incremental Learning
por: Frascaroli, Emanuele, et al.
Publicado: (2024)

Trajectory Forecasting through Low-Rank Adaptation of Discrete Latent Codes
por: Benaglia, Riccardo, et al.
Publicado: (2024)

How to Train Your Metamorphic Deep Neural Network
por: Sommariva, Thomas, et al.
Publicado: (2025)

Intrinsic Training Signals for Federated Learning Aggregation
por: Fiorini, Cosimo, et al.
Publicado: (2025)

Closed-form merging of parameter-efficient modules for Federated Continual Learning
por: Salami, Riccardo, et al.
Publicado: (2024)

May the Forgetting Be with You: Alternate Replay for Learning with Noisy Labels
por: Millunzi, Monica, et al.
Publicado: (2024)

Federated Class-Incremental Learning with Hierarchical Generative Prototypes
por: Salami, Riccardo, et al.
Publicado: (2024)

Kronecker-Factored Approximate Curvature for Physics-Informed Neural Networks
por: Dangel, Felix, et al.
Publicado: (2024)

An Attention-based Representation Distillation Baseline for Multi-Label Continual Learning
por: Menabue, Martin, et al.
Publicado: (2024)

Transporting Task Vectors across Different Architectures without Training
por: Rinaldi, Filippo, et al.
Publicado: (2026)

Kronecker-factored Approximate Curvature (KFAC) From Scratch
por: Dangel, Felix, et al.
Publicado: (2025)

Semantic Residual Prompts for Continual Learning
por: Menabue, Martin, et al.
Publicado: (2024)

Self-Labeling the Job Shop Scheduling Problem
por: Corsini, Andrea, et al.
Publicado: (2024)

Update Your Transformer to the Latest Release: Re-Basin of Task Vectors
por: Rinaldi, Filippo, et al.
Publicado: (2025)

Towards Robust Knowledge Removal in Federated Learning with High Data Heterogeneity
por: Santi, Riccardo, et al.
Publicado: (2025)

Gradient-Sign Masking for Task Vector Transport Across Pre-Trained Models
por: Rinaldi, Filippo, et al.
Publicado: (2025)

Understanding and Enforcing Weight Disentanglement in Task Arithmetic
por: Liu, Shangge, et al.
Publicado: (2026)

DOLFIN: Balancing Stability and Plasticity in Federated Continual Learning
por: Moussadek, Omayma, et al.
Publicado: (2025)

STAER: Temporal Aligned Rehearsal for Continual Spiking Neural Network
por: Gianferrari, Matteo, et al.
Publicado: (2026)

Accurate and Efficient Low-Rank Model Merging in Core Space
por: Panariello, Aniello, et al.
Publicado: (2025)

Zero-Shot Synthetic-to-Real Handwritten Text Recognition via Task Analogies
por: Garrido-Munoz, Carlos, et al.
Publicado: (2026)

Mask and Compress: Efficient Skeleton-based Action Recognition in Continual Learning
por: Mosconi, Matteo, et al.
Publicado: (2024)

Dataless Knowledge Fusion by Merging Weights of Language Models
por: Jin, Xisen, et al.
Publicado: (2022)

Generalized Fisher-Weighted SVD: Scalable Kronecker-Factored Fisher Approximation for Compressing Large Language Models
por: Chekalina, Viktoriia, et al.
Publicado: (2025)

Revisiting Scalable Hessian Diagonal Approximations for Applications in Reinforcement Learning
por: Elsayed, Mohamed, et al.
Publicado: (2024)

ABRA: Teleporting Fine-Tuned Knowledge Across Domains for Open-Vocabulary Object Detection
por: Bernardi, Mattia, et al.
Publicado: (2026)

Is Multiple Object Tracking a Matter of Specialization?
por: Mancusi, Gianluca, et al.
Publicado: (2024)

DitHub: A Modular Framework for Incremental Open-Vocabulary Object Detection
por: Cappellino, Chiara, et al.
Publicado: (2025)

Can LLMs Generate Visualizations with Dataless Prompts?
por: Coelho, Darius, et al.
Publicado: (2024)

Kronecker-Factored Approximate Curvature for Modern Neural Network Architectures
por: Eschenhagen, Runa, et al.
Publicado: (2023)

Monocular Per-Object Distance Estimation with Masked Object Modeling
por: Panariello, Aniello, et al.
Publicado: (2024)

KFCPO: Kronecker-Factored Approximated Constrained Policy Optimization
por: Lim, Joonyoung, et al.
Publicado: (2025)

CAMNet: Leveraging Cooperative Awareness Messages for Vehicle Trajectory Prediction
por: Grasselli, Mattia, et al.
Publicado: (2025)

Algebraic Priors for Approximately Equivariant Networks
por: Ali, Riccardo, et al.
Publicado: (2025)

Selective Attention-based Modulation for Continual Learning
por: Bellitto, Giovanni, et al.
Publicado: (2024)

A New Way: Kronecker-Factored Approximate Curvature Deep Hedging and its Benefits
por: Enkhbayar, Tsogt-Ochir
Publicado: (2024)