Facts as Experts: Adaptable and Interpretable Neural Memory over Symbolic Knowledge
This addresses the issue of interpretability and adaptability in large language models for NLP applications, offering a novel approach to integrate and update symbolic knowledge.
The paper tackles the problem of neural language models having inaccessible, stale, and biased factual knowledge stored in parameters by introducing a model with an explicit interface between symbolic facts and neural knowledge, which improves performance on knowledge-intensive QA tasks and allows updating facts without retraining.
Massive language models are the core of modern NLP modeling and have been shown to encode impressive amounts of commonsense and factual information. However, that knowledge exists only within the latent parameters of the model, inaccessible to inspection and interpretation, and even worse, factual information memorized from the training corpora is likely to become stale as the world changes. Knowledge stored as parameters will also inevitably exhibit all of the biases inherent in the source materials. To address these problems, we develop a neural language model that includes an explicit interface between symbolically interpretable factual information and subsymbolic neural knowledge. We show that this model dramatically improves performance on two knowledge-intensive question-answering tasks. More interestingly, the model can be updated without re-training by manipulating its symbolic representations. In particular this model allows us to add new facts and overwrite existing ones in ways that are not possible for earlier models.