Knowledge editing – Definition & Meaning

Definition of knowledge editing

: the targeted modification of facts or factual associations stored in a trained model while trying to preserve unrelated knowledge and capabilities and avoid retraining the entire model
- Knowledge editing changed the model's stored answer about the company's chief executive without noticeably affecting unrelated questions.
- The evaluation checked whether the edited fact generalized to paraphrases and whether it created contradictions elsewhere.

Origin & history

first known use2021

Knowledge editing grew into a distinct language-model research area in the early 2020s through work on locating and modifying factual associations in model parameters. The 2021 paper Editing Factual Knowledge in Language Models helped establish the modern task and its evaluation criteria.

Test yourself

Which of these is the meaning of knowledge editing?

Related terms

machine unlearning the process of removing the influence of specified training examples, knowledge, or… fine-tuning further training of an existing model on a smaller, focused dataset so it gets better at a… retrieval-augmented generation a technique in which a system first searches a collection of documents for relevant passages… training data the collection of examples (text, images, code, audio, and so on) that a model learns from hallucination a confident but false or made-up statement produced by an AI model, such as an invented fact…

Cite this entry

"Knowledge editing." AI Dictionary, Dadgogo, https://dadgogo.com/dictionary/knowledge-editing/. Accessed 8 Oct. 2026.