Data-Efficient Learning with Neural Programs
This addresses the challenge of data-efficient learning for AI systems integrating neural and symbolic components, though it appears incremental as it builds on prior gradient approximation methods.
The paper tackles the problem of learning DNN parameters in neural programs, which combine DNNs with black-box components like LLMs, using only input-output samples. The result is that their ISED algorithm achieves comparable accuracy to baselines but with improved data- and sample-efficiency, as demonstrated on benchmarks involving GPT-4 and neurosymbolic tasks.
Many computational tasks can be naturally expressed as a composition of a DNN followed by a program written in a traditional programming language or an API call to an LLM. We call such composites "neural programs" and focus on the problem of learning the DNN parameters when the training data consist of end-to-end input-output labels for the composite. When the program is written in a differentiable logic programming language, techniques from neurosymbolic learning are applicable, but in general, the learning for neural programs requires estimating the gradients of black-box components. We present an algorithm for learning neural programs, called ISED, that only relies on input-output samples of black-box components. For evaluation, we introduce new benchmarks that involve calls to modern LLMs such as GPT-4 and also consider benchmarks from the neurosymbolic learning literature. Our evaluation shows that for the latter benchmarks, ISED has comparable performance to state-of-the-art neurosymbolic frameworks. For the former, we use adaptations of prior work on gradient approximations of black-box components as a baseline, and show that ISED achieves comparable accuracy but in a more data- and sample-efficient manner.