Toward Trustworthy Neural Program Synthesis
This addresses the need for trustworthy program synthesis for developers, but it is incremental as it builds on existing language model techniques.
The paper tackles the problem of estimating the probability that a program generated by a large language model is correct, using a method that samples programs and predicates to learn a calibrated prediction model, achieving state-of-the-art generation accuracy.
We develop an approach to estimate the probability that a program sampled from a large language model is correct. Given a natural language description of a programming problem, our method samples both candidate programs as well as candidate predicates specifying how the program should behave. This allows learning a model that forms a well-calibrated probabilistic prediction of program correctness. Our system also infers which predicates are useful to explain the behavior of the generated code, and humans preferred these in a human study over raw language model outputs. Our method is simple, easy to implement, and maintains state of the art generation accuracy results.