# Ensemble Models for Neural Source Code Summarization of Subroutines

arXiv: https://arxiv.org/abs/2107.11423 · HTML: https://mlsanity.com/paper/2107.11423/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Jul 2021.

## Built on by (most cited first, 12 of 12)

- [CoditT5: Pretraining for Source Code and Natural Language Editing](https://mlsanity.com/paper/2208.05446/) (2022)
- [Multilingual training for Software Engineering](https://mlsanity.com/paper/2112.02043/) (2021)
- [A Survey on Machine Learning Techniques for Source Code Analysis](https://mlsanity.com/paper/2110.09610/) (2021)
- [Benchmarking Causal Study to Interpret Large Language Models for Source Code](https://mlsanity.com/paper/2308.12415/) (2023)
- [JAM: Controllable and Responsible Text Generation via Causal Reasoning and Latent Vector Manipulation](https://mlsanity.com/paper/2502.20684/) (2025)
- [Improving Retrieval-Augmented Code Comment Generation by Retrieving for Generation](https://mlsanity.com/paper/2408.03623/) (2024)
- [Automatic Comment Generation via Multi-Pass Deliberation](https://mlsanity.com/paper/2209.06634/) (2022)
- [Statement-based Memory for Neural Source Code Summarization](https://mlsanity.com/paper/2307.11709/) (2023)
- [Reverse Engineering User Stories from Code using Large Language Models](https://mlsanity.com/paper/2509.19587/) (2025)
- [SnipGen: A Mining Repository Framework for Evaluating LLMs for Code](https://mlsanity.com/paper/2502.07046/) (2025)
- [Meta Learning for Code Summarization](https://mlsanity.com/paper/2201.08310/) (2022)
- [Towards Comprehensive Benchmarking Infrastructure for LLMs In Software Engineering](https://mlsanity.com/paper/2601.21070/) (2026)

## Abstract (arXiv)

A source code summary of a subroutine is a brief description of that subroutine. Summaries underpin a majority of documentation consumed by programmers, such as the method summaries in JavaDocs. Source code summarization is the task of writing these summaries. At present, most state-of-the-art approaches for code summarization are neural network-based solutions akin to seq2seq, graph2seq, and other encoder-decoder architectures. The input to the encoder is source code, while the decoder helps predict the natural language summary. While these models tend to be similar in structure, evidence is emerging that different models make different contributions to prediction quality -- differences in model performance are orthogonal and complementary rather than uniform over the entire dataset. In this paper, we explore the orthogonal nature of different neural code summarization approaches and propose ensemble models to exploit this orthogonality for better overall performance. We demonstrate that a simple ensemble strategy boosts performance by up to 14.8%, and provide an explanation for this boost. The takeaway from this work is that a relatively small change to the inference procedure in most neural code summarization techniques leads to outsized improvements in prediction quality.