SemEval-2026 Task 13: Detecting Machine-Generated Code
4 views
Skip to first unread message
preslav.nakov
Nov 21, 2025, 10:36:52 AM (7 days ago) Nov 21
to SIGLEX members
SemEval-2026 Task 13: Detecting Machine-Generated Code
As AI code generators advance, distinguishing human-written code from AI-produced code is becoming critical for academic integrity, hiring, and software security.
Your challenge: Build models that can tell whether code was authored by humans or large language models (LLMs).
Subtasks:
A. Human vs. Machine: Binary classification under real-world distribution shifts.
https://www.kaggle.com/t/99673e23fe8546cf9a07a40f36f2cc7e
B. LLM Authorship Identification: Determine which LLM generated the code.
https://www.kaggle.com/t/65af9e22be6d43d884cfd6e41cad3ee4
C. Hybrid Authorship: Detect human–LLM collaboration and adversarially generated code.
https://www.kaggle.com/t/005ab8234f27424aa096b7c00a073722
Learn more: https://github.com/mbzuai-nlp/SemEval-2026-Task13
Join us in tackling one of the most pressing challenges in AI transparency and code forensics
As AI code generators advance, distinguishing human-written code from AI-produced code is becoming critical for academic integrity, hiring, and software security.
Your challenge: Build models that can tell whether code was authored by humans or large language models (LLMs).
Subtasks:
A. Human vs. Machine: Binary classification under real-world distribution shifts.
https://www.kaggle.com/t/99673e23fe8546cf9a07a40f36f2cc7e
B. LLM Authorship Identification: Determine which LLM generated the code.
https://www.kaggle.com/t/65af9e22be6d43d884cfd6e41cad3ee4
C. Hybrid Authorship: Detect human–LLM collaboration and adversarially generated code.
https://www.kaggle.com/t/005ab8234f27424aa096b7c00a073722
Learn more: https://github.com/mbzuai-nlp/SemEval-2026-Task13
Join us in tackling one of the most pressing challenges in AI transparency and code forensics
Reply all
Reply to author
Forward
0 new messages