Papers with Code 아카이브
Task Arithmetic
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 하위 태스크
- 0
- 평가표
- 0
- 보존된 평가 기록
- 0
A task vector specifies a direction in the weight space of a pre-trained model, such that movement in that direction improves performance on the task. We build task vectors by subtracting the weights of a pre-trained model from the weights of the same model after fine-tuning on a task. We show that these task vectors can be modified and combined together through arithmetic operations such as negation and addition, and the behavior of the resulting model is steered accordingly.
논문
이 태스크와 하위 분야의 관련 논문입니다. 실제 성능 비교는 평가 결과에서 확인하세요.
DuET: Dual Incremental Object Detection via Exemplar-Free Task Arithmetic
CultureMERT: Continual Pre-Training for Cross-Cultural Music Representation Learning
Subspace-Boosted Model Merging
CALM: Consensus-Aware Localized Merging for Multi-Task Learning
FedRPCA: Enhancing Federated LoRA Aggregation Using Robust PCA
On Fairness of Task Arithmetic: The Role of Task Vectors
Scalable Strategies for Continual Learning with Replay
Cross-Model Transfer of Task Vectors via Few-Shot Orthogonal Alignment
MCU: Improving Machine Unlearning through Mode Connectivity
CAT Merging: A Training-Free Approach for Resolving Conflicts in Model Merging
Investigating Task Arithmetic for Zero-Shot Information Retrieval
Leveraging Submodule Linearity Enhances Task Arithmetic Performance in LLMs
When is Task Vector Provably Effective for Model Editing? A Generalization Analysis of Nonlinear Transformers
Single-Input Multi-Output Model Merging: Leveraging Foundation Models for Dense Multi-Task Learning
Efficient Model Editing with Task-Localized Sparse Fine-tuning
OpenThaiGPT 1.6 and R1: Thai-Centric Open Source and Reasoning Large Language Models
Disentangling Task Interference within Neurons: Model Merging in Alignment with Neuronal Mechanisms
Layer-Aware Task Arithmetic: Disentangling Task-Specific and Instruction-Following Knowledge
Neural Networks Remember More: The Power of Parameter Isolation and Combination
Mediator: Memory-efficient LLM Merging with Less Parameter Conflicts and Uncertainty Based Routing
Efficient Model Editing with Task Vector Bases: A Theoretical Framework and Scalable Approach
Task Arithmetic in Trust Region: A Training-Free Model Merging Approach to Navigate Knowledge Conflicts
Soup to go: mitigating forgetting during continual learning with model averaging
BADTV: Unveiling Backdoor Threats in Third-Party Task Vectors
Bias Vector: Mitigating Biases in Language Models with Task Arithmetic Approach
Task Arithmetic Through The Lens Of One-Shot Federated Learning
Multi-Task Model Merging via Adaptive Weight Disentanglement
Task Singular Vectors: Reducing Task Interference in Model Merging
Beyond Task Vectors: Selective Task Arithmetic Based on Importance Metrics
ATM: Improving Model Merging by Alternating Tuning and Merging
Efficient and Effective Weight-Ensembling Mixture of Experts for Multi-Task Model Merging
The Non-Local Model Merging Problem: Permutation Symmetries and Variance Collapse
NegMerge: Consensual Weight Negation for Strong Machine Unlearning
What Matters for Model Merging at Scale?
Towards Diverse Device Heterogeneous Federated Learning via Task Arithmetic Knowledge Integration
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.