model editing
Model editing is the process of making specific, targeted adjustments to an AI model's behaviors or outputs without retraining from scratch. This can include altering training data, modifying model parameters, or refining structural components.
- A$^3$E: Towards Compositional Model Editing
- Conflict-Aware Knowledge Editing in the Wild: Semantic-Augmented Graph Representation for Unstructured Text
- Exploring and Leveraging Class Vectors for Classifier Editing
- Hippocampal-like Sequential Editing for Continual Knowledge Updates in Large Language Models
- Localizing Knowledge in Diffusion Transformers
- Model Editing for Vision Transformers
- Rethinking Residual Distribution in Locate-then-Edit Model Editing
- UniEdit: A Unified Knowledge Editing Benchmark for Large Language Models