The Dagstuhl Seminar 24431 on "Automated Programming and Program Repair" brought together 33 researchers from academia and industry to explore the intersection of automated code generation and program repair. Over five days (October 21–25, 2024), participants discussed advances in large language models (LLMs) for code generation, the role of automated program repair in improving generated code, and challenges in deploying these technologies in real-world software development. The seminar featured over 20 talks and three panel discussions on topics such as benchmarks for LLM-generated code, trust in automated programming, and the broader applications of LLMs beyond coding assistance. Key outcomes included identifying critical challenges in benchmarking, evaluation criteria, and developer adoption of automated repair techniques, fostering future collaborations and actionable research directions in the field.
@Article{legoues_et_al:DagRep.14.10.39, author = {Le Goues, Claire and Pradel, Michael and Roychoudhury, Abhik and Tan, Shin Hwei}, title = {{Automated Programming and Program Repair (Dagstuhl Seminar 24431)}}, pages = {39--57}, journal = {Dagstuhl Reports}, ISSN = {2192-5283}, year = {2025}, volume = {14}, number = {10}, editor = {Le Goues, Claire and Pradel, Michael and Roychoudhury, Abhik and Tan, Shin Hwei}, publisher = {Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik}, address = {Dagstuhl, Germany}, URL = {https://drops.dagstuhl.de/entities/document/10.4230/DagRep.14.10.39}, URN = {urn:nbn:de:0030-drops-230235}, doi = {10.4230/DagRep.14.10.39}, annote = {Keywords: Auto-coding, Large Language Models, Automated Program Repair, Program Synthesis, Trustworthy Software} }
Feedback for Dagstuhl Publishing