First Workshop on Large Language Models For Generative Software Engineering (LLM4SE 2025)
The workshop is co-located with Software Technologies: Applications and Foundations (STAF2025)

website: https://conf.researchr.org/home/staf-2025/llm4se-2025

LLMs tailored for code, such as CodeLLama, Codex, and others, have revolutionised the landscape of software engineering. They significantly enhance developers’ efficiency in several classical software engineering tasks like coding, testing, and documentation. LLMs are modifying how we interact with computers, and their application to software engineering holds immense potential. This workshop will explore how these powerful models can be harnessed to automate, augment, and even transform the software development lifecycle, leading to increased productivity, innovative solutions, and a new era of software creation.

Hence, the primary objective of this workshop is to explore potential applications where LLMs can support software engineers and engage in discussions regarding the implementation of such solutions. As the workshop is a direct evolution of the previous edition named “Large Language Models for Model Driven Engineering”, we consider Model-Driven Engineering (MDE) as a cornerstone of this workshop.

The objectives of the llm4se 2025 workshop are to bring together researchers, practitioners and tool developers working on applying Large Language Models to tackle software engineering tasks. We aim to foster discussions and reflections on the benefits and threats of using Large Language Models.

-----------------------
Topic of Interest
-----------------------

Topics of interest for the workshop include, but are not limited to:

- LLMs for Requirements Generation (e.g. requirements collection, organisation, analysis)
- LLMs for Verification and Validation (e.g. test case generation, model-based testing, functional testing)
- LLMs for Model Generation
- LLMs for Low-Code Generation
- LLMs for Code Generation
- LLMs for Code Summarization (e.g. description generation, comments generation)
- LLMs for Patch Generation (e.g. repair, bugs detection)
- LLMs for Code Optimization (e.g. performance, energy)
- LLMs for Code Translation

Submission Info

Two kinds of papers are solicited:

- regular papers (10 pages + 1 page for references)
- short papers (4 pages + 1 page for references)

Contributions should present novel research ideas (even if at a preliminary development stage), challenging problems, and practical contributions to the domain. Industrial experience reports or case studies related to the development or use of LLMs in a broad software engineering context are also solicited. Furthermore, extended abstract are accepted, that can include prior work or visionary work, but will not included in the proceedings. All papers must be written in English adhere to the CEUR Style single-column formatting guidelines, and be submitted through EasyChair at the link: https://easychair.org/conferences/?conf=staf2025 and then selecting the track Large Language Models for Software Engineering (LLM4SE) .

-----------------------
Important Dates
-----------------------

- Abstract Submission: March 28th, 2025 AoE
- Paper Submission: April 4th, 2025 AoE
- Author Notification: May 5th, 2025 AoE
- Camera-Ready: May 19th, 2025 AoE

----------------
Organizers
----------------

Riccardo Rubei (University of L’Aquila, Italy)
José Antonio Hernandéz López (Linköping University, Sweden)
Antonio Cicchetti (Mälardalens University, Sweden)