K-Edit: Language Model Editing with Contextual Knowledge Awareness
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913709620723712 |
|---|---|
| author | Markowitz, Elan Ramakrishna, Anil Mehrabi, Ninareh Peris, Charith Gupta, Rahul Chang, Kai-Wei Galstyan, Aram |
| author_facet | Markowitz, Elan Ramakrishna, Anil Mehrabi, Ninareh Peris, Charith Gupta, Rahul Chang, Kai-Wei Galstyan, Aram |
| contents | As the world changes, we need to be able to update our models and correct false information without costly retraining. Knowledge-based model editing enables precise modifications to the weights of large language models in order to modify the information encoded within. Recent approaches have seen success in enabling recall of edited information for thousands of edits at once. However, these approaches fail to produce edits that account for associated contextual information. We present K-Edit, an effective approach to generating contextually consistent knowledge edits. By using knowledge graphs, which maintain contextual consistency when an edge is edited, we are able to generate additional \textit{contextual edits} that ensure consistency of related information in the language model. Our experiments demonstrate significant improvements in multi-hop question answering while maintaining the general effectiveness and scalability of model edits. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_10626 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | K-Edit: Language Model Editing with Contextual Knowledge Awareness Markowitz, Elan Ramakrishna, Anil Mehrabi, Ninareh Peris, Charith Gupta, Rahul Chang, Kai-Wei Galstyan, Aram Machine Learning Artificial Intelligence As the world changes, we need to be able to update our models and correct false information without costly retraining. Knowledge-based model editing enables precise modifications to the weights of large language models in order to modify the information encoded within. Recent approaches have seen success in enabling recall of edited information for thousands of edits at once. However, these approaches fail to produce edits that account for associated contextual information. We present K-Edit, an effective approach to generating contextually consistent knowledge edits. By using knowledge graphs, which maintain contextual consistency when an edge is edited, we are able to generate additional \textit{contextual edits} that ensure consistency of related information in the language model. Our experiments demonstrate significant improvements in multi-hop question answering while maintaining the general effectiveness and scalability of model edits. |
| title | K-Edit: Language Model Editing with Contextual Knowledge Awareness |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2502.10626 |