Is Programming by Example solved by LLMs?

TL;DR

This paper evaluates the capability of Large Language Models in solving Programming-by-Examples tasks, revealing they excel with fine-tuning on in-distribution data but struggle with out-of-distribution problems, highlighting both progress and limitations.

Contribution

The study provides empirical analysis of LLMs on PBE tasks across domains, demonstrating the importance of fine-tuning and offering insights into generalization challenges.

Findings

Pretrained LLMs are ineffective at PBE tasks without fine-tuning.

Fine-tuning significantly improves performance on in-distribution problems.

Out-of-distribution generalization remains a key challenge for LLMs in PBE.

Abstract

Programming-by-Examples (PBE) aims to generate an algorithm from input-output examples. Such systems are practically and theoretically important: from an end-user perspective, they are deployed to millions of people, and from an AI perspective, PBE corresponds to a very general form of few-shot inductive inference. Given the success of Large Language Models (LLMs) in code-generation tasks, we investigate here the extent to which LLMs can be said to have "solved" PBE. We experiment on classic domains such as lists and strings, and an uncommon graphics programming domain not well represented in typical pretraining data. We find that pretrained models are not effective at PBE, but that they can be fine-tuned for much higher performance, provided the test problems are in-distribution. We analyze empirically what causes these models to succeed and fail, and take steps toward understanding how to achieve better out-of-distribution generalization. Collectively these results suggest that LLMs make strong progress toward solving the typical suite of PBE tasks, potentially increasing the flexibility and applicability of PBE systems, while also identifying ways in which LLMs still fall short.