Note that the terminology used in PCCoder differs from what we have used here. The inputs to a program can either be an integer or an array of integers of maximum length 20. There can be a maximum of three input arguments to a program. There are 1298 statements and 38 operators in the DSL, i.e., This forms the state X . 1 State Embedding ( H For the next steps, we repeat the last two steps mentioned above till we reach the end of the program. See Figure 2 for an illustration of the process at t = 2 .

These systems try to find a program that satisfies all IO examples simultaneously, which under most of the settings can be hard. What if instead, we try to find this program in parts?

The goal of program synthesis from examples is to find a computer program that is consistent with a given set of input-output examples. Most learning-based approaches try to find a program that satisfies all examples at once. Our work, by contrast, considers an approach that breaks the problem into two stages: (a) find programs that satisfy only one example, and (b) leverage these per-example solutions to yield a program that satisfies all examples. We introduce the Cross Aggregator neural network module based on a multi-head attention mechanism that learns to combine the cues present in these per-example solutions to synthesize a global solution. Evaluation across programs of different lengths and under two different experimental settings reveal that when given the same time budget, our technique significantly improves the success rate over PCCoder [32] and other ablation baselines.