Neural Algorithmic Reasoning Without Intermediate Supervision
This work addresses the challenge of generalizing neural algorithmic reasoning to out-of-distribution data, such as larger input sizes, without relying on intermediate step supervision, which is incremental but promising for improving efficiency in algorithm imitation tasks.
The paper tackles the problem of learning neural algorithmic reasoning without intermediate supervision, proposing architectural improvements and a self-supervised objective to regularize intermediate computations, achieving competitive results and new state-of-the-art performance on tasks like sorting with significant improvements.
Neural algorithmic reasoning is an emerging area of machine learning focusing on building models that can imitate the execution of classic algorithms, such as sorting, shortest paths, etc. One of the main challenges is to learn algorithms that are able to generalize to out-of-distribution data, in particular with significantly larger input sizes. Recent work on this problem has demonstrated the advantages of learning algorithms step-by-step, giving models access to all intermediate steps of the original algorithm. In this work, we instead focus on learning neural algorithmic reasoning only from the input-output pairs without appealing to the intermediate supervision. We propose simple but effective architectural improvements and also build a self-supervised objective that can regularise intermediate computations of the model without access to the algorithm trajectory. We demonstrate that our approach is competitive to its trajectory-supervised counterpart on tasks from the CLRS Algorithmic Reasoning Benchmark and achieves new state-of-the-art results for several problems, including sorting, where we obtain significant improvements. Thus, learning without intermediate supervision is a promising direction for further research on neural reasoners.