Making Neural Programming Architectures Generalize via Recursion
This addresses the challenge of making neural programming architectures more robust and interpretable for AI researchers and practitioners, though it is incremental as it builds on existing frameworks like Neural Programmer-Interpreter.
The paper tackled the problem of poor generalizability and limited interpretability in neural networks learning programs by augmenting architectures with recursion, demonstrating superior generalizability and interpretability on tasks like grade-school addition and sorting with small training data.
Empirically, neural networks that attempt to learn programs from data have exhibited poor generalizability. Moreover, it has traditionally been difficult to reason about the behavior of these models beyond a certain level of input complexity. In order to address these issues, we propose augmenting neural architectures with a key abstraction: recursion. As an application, we implement recursion in the Neural Programmer-Interpreter framework on four tasks: grade-school addition, bubble sort, topological sort, and quicksort. We demonstrate superior generalizability and interpretability with small amounts of training data. Recursion divides the problem into smaller pieces and drastically reduces the domain of each neural network component, making it tractable to prove guarantees about the overall system's behavior. Our experience suggests that in order for neural architectures to robustly learn program semantics, it is necessary to incorporate a concept like recursion.