No Large Language Models (LLMs): Use of any LLMs—including open-source models, API-based models, foundation models, or instruction-tuned transformers—is strictly prohibited across all tracks.

Disqualification: Any violation of the LLM policy or other rules will result in immediate disqualification.

Authorized Data Only: Participants may only use the datasets explicitly provided for their specific track.

No External Data: The use of external datasets is strictly prohibited.

Reproducibility: All work must be fully reproducible and clearly documented to be eligible for judging.

Track Participation: Participants may compete in one or more tracks, provided they follow the specific datasets and rules for each.

Original Work: Projects must be created specifically for this event; using prior work is not allowed.

Team Participation: Participants may only contribute to one team per event.