Logical Interpretations of Autoencoders
This work addresses the challenge of unifying perception and reasoning in AI, offering a novel approach for interpretability in deep learning, though it appears incremental in the broader context of integrating symbolic and neural methods.
The paper tackles the problem of interpreting what hidden layers in autoencoders capture by integrating a symbolic generative framework onto the feature layer, enabling logical reasoning and tasks like generating example images for classes and handling noisy labels, with empirical evaluations showing promise.
The unification of low-level perception and high-level reasoning is a long-standing problem in artificial intelligence, which has the potential to not only bring the areas of logic and learning closer together but also demonstrate how abstract concepts might emerge from sensory data. Precisely because deep learning methods dominate perception-based learning, including vision, speech, and linguistic grammar, there is fast-growing literature on how to integrate symbolic reasoning and deep learning. Broadly, efforts seem to fall into three camps: those focused on defining a logic whose formulas capture deep learning, ones that integrate symbolic constraints in deep learning, and others that allow neural computations and symbolic reasoning to co-exist separately, to enjoy the strengths of both worlds. In this paper, we identify another dimension to this inquiry: what do the hidden layers really capture, and how can we reason about that logically? In particular, we consider autoencoders that are widely used for dimensionality reduction and inject a symbolic generative framework onto the feature layer. This allows us, among other things, to generate example images for a class to get a sense of what was learned. Moreover, the modular structure of the proposed model makes it possible to learn relations over multiple images at a time, as well as handle noisy labels. Our empirical evaluations show the promise of this inquiry.