%0 Conference Proceedings %B Neural Information Processing Systems (NeurIPS 2019) %D 2019 %T Write, Execute, Assess: Program Synthesis with a REPL %A Kevin Ellis %A Maxwell Nye %A Yewen Pu %A Felix Sosa %A Joshua B. Tenenbaum %A Armando Solar-Lezama %X

We present a neural program synthesis approach integrating components which write, execute, and assess code to navigate the search space of possible programs. We equip the search process with an interpreter or a read-eval-print-loop (REPL), which immediately executes partially written programs, exposing their semantics. The REPL addresses a basic challenge of program synthesis: tiny changes in syntax can lead to huge changes in semantics. We train a pair of models, a policy that proposes the new piece of code to write, and a value function that assesses the prospects of the code written so-far. At test time we can combine these models with a Sequential Monte Carlo algorithm. We apply our approach to two domains: synthesizing text editing programs and inferring 2D and 3D graphics programs.

%B Neural Information Processing Systems (NeurIPS 2019) %C Vancouver, Canada %8 11/2019 %G eng