Copertina

Decomposing Task Vectors for Refined Model Editing

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Damirchi, Hamed, Abbasnejad, Ehsan, Zhang, Zhen, Shi, Javen
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

The Quest for Winning Tickets in Low-Rank Adapters
di: Damirchi, Hamed, et al.
Pubblicazione: (2025)

Truth as a Trajectory: What Internal Representations Reveal About Large Language Model Reasoning
di: Damirchi, Hamed, et al.
Pubblicazione: (2026)

Learning Latent Dynamical Causal Processes for Single-Cell Perturbation Prediction
di: Jiang, Wenkang, et al.
Pubblicazione: (2026)

Synergy and Diversity in CLIP: Enhancing Performance Through Adaptive Backbone Ensembling
di: Rodriguez-Opazo, Cristian, et al.
Pubblicazione: (2024)

What Makes a Representation Good for Single-Cell Perturbation Prediction?
di: Jiang, Wenkang, et al.
Pubblicazione: (2026)

Knowledge Composition using Task Vectors with Learned Anisotropic Scaling
di: Zhang, Frederic Z., et al.
Pubblicazione: (2024)

Rethinking State Disentanglement in Causal Reinforcement Learning
di: Cao, Haiyao, et al.
Pubblicazione: (2024)

Decomposing and Editing Predictions by Modeling Model Computation
di: Shah, Harshay, et al.
Pubblicazione: (2024)

BruSLeAttack: A Query-Efficient Score-Based Black-Box Sparse Adversarial Attack
di: Vo, Viet Quoc, et al.
Pubblicazione: (2024)

When is Task Vector Provably Effective for Model Editing? A Generalization Analysis of Nonlinear Transformers
di: Li, Hongkang, et al.
Pubblicazione: (2025)

Beyond Imitation: Recovering Dense Rewards from Demonstrations
di: Li, Jiangnan, et al.
Pubblicazione: (2025)

Do Deep Neural Network Solutions Form a Star Domain?
di: Sonthalia, Ankit, et al.
Pubblicazione: (2024)

Highway Graph to Accelerate Reinforcement Learning
di: Yin, Zidu, et al.
Pubblicazione: (2024)

Do We Always Need the Simplicity Bias? Looking for Optimal Inductive Biases in the Wild
di: Teney, Damien, et al.
Pubblicazione: (2025)

Exploring Context Window of Large Language Models via Decomposed Positional Vectors
di: Dong, Zican, et al.
Pubblicazione: (2024)

Variational Task Vector Composition
di: Zhang, Boyuan, et al.
Pubblicazione: (2025)

Neural Redshift: Random Networks are not Random Functions
di: Teney, Damien, et al.
Pubblicazione: (2024)

A Survey on Deep Neural Network Pruning-Taxonomy, Comparison, Analysis, and Recommendations
di: Cheng, Hongrong, et al.
Pubblicazione: (2023)

ETAGE: Enhanced Test Time Adaptation with Integrated Entropy and Gradient Norms for Robust Model Performance
di: Shamsi, Afshar, et al.
Pubblicazione: (2024)

Identifying Weight-Variant Latent Causal Models
di: Liu, Yuhang, et al.
Pubblicazione: (2022)

Premonition: Using Generative Models to Preempt Future Data Changes in Continual Learning
di: McDonnell, Mark D., et al.
Pubblicazione: (2024)

Identifiable Latent Polynomial Causal Models Through the Lens of Change
di: Liu, Yuhang, et al.
Pubblicazione: (2023)

RanPAC: Random Projections and Pre-trained Models for Continual Learning
di: McDonnell, Mark D., et al.
Pubblicazione: (2023)

CEDL: Centre-Enhanced Discriminative Learning for Anomaly Detection
di: Darban, Zahra Zamanzadeh, et al.
Pubblicazione: (2025)

Towards Identifiable Latent Additive Noise Models
di: Liu, Yuhang, et al.
Pubblicazione: (2024)

Chem4DLLM: 4D Multimodal LLMs for Chemical Dynamics Understanding
di: Li, Xinyu, et al.
Pubblicazione: (2026)

Latent Covariate Shift: Unlocking Partial Identifiability for Multi-Source Domain Adaptation
di: Liu, Yuhang, et al.
Pubblicazione: (2022)

VectorEdits: A Dataset and Benchmark for Instruction-Based Editing of Vector Graphics
di: Kuchař, Josef, et al.
Pubblicazione: (2025)

The Geometric Mechanics of Contrastive Representation Learning: Alignment Potentials, Entropic Dispersion, and Cross-modal Divergence
di: Cai, Yichao, et al.
Pubblicazione: (2026)

Towards Higher Effective Rank in Parameter-efficient Fine-tuning using Khatri--Rao Product
di: Albert, Paul, et al.
Pubblicazione: (2025)

On Fairness of Task Arithmetic: The Role of Task Vectors
di: Naganuma, Hiroki, et al.
Pubblicazione: (2025)

Task Vector Quantization for Memory-Efficient Model Merging
di: Kim, Youngeun, et al.
Pubblicazione: (2025)

Analytic DAG Constraints for Differentiable DAG Learning
di: Zhang, Zhen, et al.
Pubblicazione: (2025)

The Character Error Vector: Decomposable errors for page-level OCR evaluation
di: Bourne, Jonathan, et al.
Pubblicazione: (2026)

Concept Component Analysis: A Principled Approach for Concept Extraction in LLMs
di: Liu, Yuhang, et al.
Pubblicazione: (2026)

Adaptive Task Vectors for Large Language Models
di: Kang, Joonseong, et al.
Pubblicazione: (2025)

Decomposed Inductive Procedure Learning: Learning Academic Tasks with Human-Like Data Efficiency
di: Weitekamp, Daniel, et al.
Pubblicazione: (2025)

On Task Vectors and Gradients
di: Zhou, Luca, et al.
Pubblicazione: (2025)

Beyond DAGs: A Latent Partial Causal Model for Multimodal Learning
di: Liu, Yuhang, et al.
Pubblicazione: (2024)

Policy Adaptation via Language Optimization: Decomposing Tasks for Few-Shot Imitation
di: Myers, Vivek, et al.
Pubblicazione: (2024)