Learning Task Decomposition to Assist Humans in Competitive Programming
This addresses the challenge for humans in repairing complex AI-generated solutions, offering a practical assistive tool with incremental improvements in human-AI collaboration.
The paper tackles the problem of humans struggling to understand and repair language model-generated solutions by proposing automatic task decomposition into simpler subtasks, resulting in non-experts solving 33.3% more problems, speeding up by 3.3x, and matching unassisted experts in competitive programming.
When using language models (LMs) to solve complex problems, humans might struggle to understand the LM-generated solutions and repair the flawed ones. To assist humans in repairing them, we propose to automatically decompose complex solutions into multiple simpler pieces that correspond to specific subtasks. We introduce a novel objective for learning task decomposition, termed assistive value (AssistV), which measures the feasibility and speed for humans to repair the decomposed solution. We collect a dataset of human repair experiences on different decomposed solutions. Utilizing the collected data as in-context examples, we then learn to critique, refine, and rank decomposed solutions to improve AssistV. We validate our method under competitive programming problems: under 177 hours of human study, our method enables non-experts to solve 33.3\% more problems, speeds them up by 3.3x, and empowers them to match unassisted experts.