Join the Data Contamination Shared Task at CONDA 2024 Workshop: Gather Evidence and Contribute to Fair Evaluation

The CONDA 2024 workshop invites the NLP community to participate in a shared task focused on data contamination evidence collection. With growing concerns over the impact of contaminated pre-training data on language models, this shared task aims to provide a centralized platform for evidence submission and understanding the extent of the problem. Participants will report cases of contamination or lack thereof, either in pre-training corpora or models, and contribute to a compilation paper. Submissions must follow the provided guidelines and will be reviewed by the organizers. Evidence submission deadline is July 1, 2024, and the workshop day is scheduled for August 16, 2024. Sponsors include AWS, HuggingFace, and Google.