Efficiency and safety of Large Language Models (LLMs), among other factors, rely on the quality of tokenization.A good tokenizer not only improves inference speed and language understanding but also provides extra defense against jailbreak attacks and lowers the risk of hallucinations.In this work, we investigate the efficiency of code tokenization, in particular from the perspective of data source diversity.We demonstrate that code tokenizers are prone to producing unused, and thus under-trained, tokens due to the imbalance in repository and language diversity in the training data, as well as the dominance of source-specific, repetitive tokens that are often unusable in future inference.By modifying the BPE objective and introducing merge skipping, we implement different techniques under the name Source-Attributed BPE (SA-BPE) to regularize BPE training and minimize overfitting, thereby substantially reducing the number of under-trained tokens while maintaining the same inference procedure as with regular BPE.This provides an effective tool suitable for production use. pchizhov/sa-bpe

Token-Level code completion is one of the most critical features in modern Integrated Development Environments (IDEs). It assists developers by suggesting relevant identifiers and APIs during coding. While completions are typically derived from static analysis, their usefulness depends heavily on how they are ranked, as correct predictions buried deep in the list are rarely seen by users. Most current systems rely on handcrafted heuristics or lightweight machine learning models trained on user logs, which can be further improved to capture context information and generalize across projects and coding styles. In this work, we propose a new scoring approach to ranking static completions using language models in a lightweight and model-agnostic way. Our method organizes all valid completions into a prefix tree and performs a single greedy decoding pass to collect token-level scores across the tree. This enables a precise token-aware ranking without needing beam search, prompt engineering, or model adaptations. The approach is fast, architecture-agnostic, and compatible with already deployed models for code completion. These findings highlight a practical and effective pathway for integrating language models into already existing tools within IDEs, and ultimately providing smarter and more responsive developer assistance.

This paper introduces the human-curated Pandas-PlotBench dataset, designed to evaluate language models’ effectiveness as assistants in visual data exploration. Our benchmark focuses on generating code for visualizing tabular data—such as a Pandas DataFrame—based on natural language instructions, complementing current evaluation tools and expanding their scope. The dataset includes 175 unique tasks. Our experiments assess several leading Large Language Models (LLMs) across three visualization libraries: Matplotlib, Seaborn, and Plotly. We show that the shortening of tasks has a minimal effect on plotting capabilities, allowing for the user interface that accommodates concise user input without sacrificing functionality or accuracy. Another of our findings reveals that while LLMs perform well with popular libraries like Matplotlib and Seaborn, challenges persist with Plotly, highlighting areas for improvement. We hope that the modular design of our benchmark will broaden the current studies on generating visualizations. Our dataset and benchmark code is available online: https://huggingface. co/datasets/JetBrains-Research/PandasPlotBench; https://github.com/JetBrains-Research/PandasPlotBench.

Long methods that encapsulate multiple responsibilities within a single method are challenging to maintain. Choosing which statements to extract into new methods has been the target of many research tools. Despite steady improvements, these tools often fail to generate refactorings that align with developers' preferences and acceptance criteria. Given that Large Language Models (LLMs) have been trained on large code corpora, if we harness their familiarity with the way developers form functions, we could suggest refactorings that developers are likely to accept. In this paper, we advance the science and practice of refactoring by synergistically combining the insights of LLMs with the power of IDEs to perform Extract Method (EM). Our formative study on 1752 EM scenarios revealed that LLMs are very effective for giving expert suggestions, yet they are unreliable: up to 76.3% of the suggestions are hallucinations. We designed a novel approach that removes hallucinations from the candidates suggested by LLMs, then further enhances and ranks suggestions based on static analysis techniques from program slicing, and finally leverages the IDE to execute refactorings correctly. We implemented this approach in an IntelliJ IDEA plugin called EM-Assist. We empirically evaluated EM-Assist on a diverse corpus that replicates 1752 actual refactorings from open-source projects. We found that EM-Assist outperforms previous state of the art tools: EM-Assist suggests the developer-performed refactoring in 53.4% of cases, improving over the recall rate of 39.4% for previous best-in-class tools. Furthermore, we conducted firehouse surveys with 16 industrial developers and suggested refactorings on their recent commits. 81.3% of them agreed with the recommendations provided by EM-Assist. This shows the usefulness of our approach and ushers us into a new era when LLMs become effective AI assistants for refactoring.

Commit messages are crucial to software development, allowing developers to track changes and collaborate effectively. Despite their utility, most commit messages lack important information since writing high-quality commit messages is tedious and time-consuming. The active research on commit message generation (CMG) has not yet led to wide adoption in practice. We argue that if we could shift the focus from commit message generation to commit message completion and use previous commit history as additional context, we could significantly improve the quality and the personal nature of the resulting commit messages. In this paper, we propose and evaluate both of these novel ideas. Since the existing datasets lack historical data, we collect and share a novel dataset called CommitChronicle, containing 10.7M commits across 20 programming languages. We use this dataset to evaluate the completion setting and the usefulness of the historical context for state-of-the-art CMG models and GPT-3.5-turbo. Our results show that in some contexts, commit message completion shows better results than generation, and that while in general GPT-3.5-turbo performs worse, it shows potential for long and detailed messages. As for the history, the results show that historical information improves the performance of CMG models in the generation task, and the performance of GPT-3.5-turbo in both generation and completion.

Authorship attribution (i.e., determining who is the author of a piece of source code) is an established research topic. State-of-the-art results for the authorship attribution problem look promising for the software engineering field, where they could be applied to detect plagiarized code and prevent legal issues. With this article, we first introduce a new language-agnostic approach to authorship attribution of source code. Then, we discuss limitations of existing synthetic datasets for authorship attribution, and propose a data collection approach that delivers datasets that better reflect aspects important for potential practical use in software engineering. Finally, we demonstrate that high accuracy of authorship attribution models on existing datasets drastically drops when they are evaluated on more realistic data. We outline next steps for the design and evaluation of authorship attribution models that could bring the research efforts closer to practical use for software engineering.

In this paper, we present Sosed, a tool for discovering similar software projects. We use fastText to compute the embeddings of subtokens into a dense space for 120,000 GitHub projects in 200 languages. Then, we cluster embeddings to identify groups of semantically similar subtokens that reflect topics in source code. We use a dataset of 9 million GitHub projects as a reference search base. To identify similar projects, we compare the distributions of clusters among their subtokens. The tool receives an arbitrary project as input, extracts subtokens in 16 most popular programming languages, computes cluster distribution, and finds projects with the closest distribution in the search base. We labeled subtoken clusters with short descriptions to enable Sosed to produce interpretable output.

One recent, significant advance in modeling source code for machine learning algorithms has been the introduction of path-based representation - an approach consisting in representing a snippet of code as a collection of paths from its syntax tree. Such representation efficiently captures the structure of code, which, in turn, carries its semantics and other information. Building the path-based representation involves parsing the code and extracting the paths from its syntax tree; these steps build up to a substantial technical job. With no common reusable toolkit existing for this task, the burden of mining diverts the focus of researchers from the essential work and hinders newcomers in the field of machine learning on code. In this paper, we present PathMiner - an open-source library for mining path-based representations of code. PathMiner is fast, flexible, well-tested, and easily extensible to support input code in any common programming language. Preprint [https://doi.org/10.5281/zenodo.2595271]; released tool [https://doi.org/10.5281/zenodo.2595257].