Contextualized Code Pretraining for Code Generation

TL;DR

This paper introduces a new invocation-aware pretraining framework for code models that leverages calling context to improve code generation in realistic repository scenarios.

Contribution

It proposes contextualized code pretraining using static analysis to incorporate calling context, and introduces CallerGen models and CallerEval benchmark.

Findings

CallerGen outperforms comparable models on CallerEval

Models trained with invocation-aware objectives achieve higher pass@1 scores

Calling context significantly improves code generation performance

Abstract

As code generation becomes increasingly central to improving software development efficiency, modern code models are largely trained and evaluated on code with natural-language descriptions. In real projects, developers often implement missing functions under limited project-specific artifacts, while the local call-site context is already available in the surrounding code. This usage context provides actionable cues about expected behavior, but existing models are not explicitly optimized to leverage it reliably, leading to implementations that may not integrate smoothly with surrounding usage in repository settings. In this work, we propose contextualized code pretraining, an invocation-aware framework that integrates calling context into both the training and evaluation of code models. Using static analysis, we automatically extract large-scale caller-callee pairs from real repositories to construct pretraining tasks and benchmarks that condition generation on the calling context. We train CallerGen, the first code models pretrained with invocation-aware objectives spanning multiple sizes, and evaluate them on CallerEval, a new benchmark featuring realistic scenarios. Experiments show that CallerGen outperforms comparable-scale models and remains competitive with larger ones across two benchmarks. Our 220M and 0.5B models achieve 16.58% and 22.81@% pass1, surpassing baselines on CallerEval. These results highlight the importance of calling context in realistic code generation.