Semi-Parametric Model Editor with a Retrieval-Augmented Counterfactual Model

Home

Discover

Bookmarks

Keep Up With Latest Trending Papers. Computer Science, AI and Machine Learning and more.Subscribe

CC BY

Source arXiv

Artificial Intelligence

Computation and Language

Semi-Parametric Model Editor with a Retrieval-Augmented Counterfactual Model

Memory-Based Model Editing at Scale

Model editors make local updates to the behavior of base (pre-trained) models to inject updated knowledge or correct undesirable behaviors.

Existing approaches to model editing have shown promise, but also suffer from insufficient expressiveness : they struggle to accurately model an edit s intended scope (examples affected by the edit), leading to inaccuratepredictions for test inputs loosely related to the edit, and they often fail altogether after many edits.

As a higher-capacity alternative, we propose semi-parametric editing with a retrieval-augmented counterfactual model(serac), which stores edits in an explicit memory and learns to reason over them to modulate the base model s predictions as needed.

We introduce three challenging language model editing problems based on question answering, fact-checking, and dialogue generation.

We find that only serac achieves high performance on all three problems, consistently outperforming existing approaches to model editing by a significant margin.

Authors

Eric Mitchell, Charles Lin, Antoine Bosselut, Christopher D. Manning, Chelsea Finn

◐ Recommended Follows

◐ Latest Activity