SHROOM-CAP Shared Task: Hallucination Detection for Scientific Content
Participate in the SHROOM-CAP shared task to detect hallucinations in scientific content generated by LLMs across multiple languages.
The SHROOM-CAP shared task is an Indic-centric initiative co-located with the CHOMPS-2025 workshop at IJCNLP-AACL 2025, focused on advancing the state-of-the-art in hallucination detection for scientific content generated by Large Language Models (LLMs).
Key highlights include:
- LLM-centered approach
- Cross-lingual annotations
- Hallucination and fluency prediction
Participants are invited to detect hallucinations in scientific content across multiple languages, including English, French, Spanish, Hindi, Italian, and several Indic languages.
To participate, register your team and join the SHROOM-CAP Google group. Submit your results by 12.10.2025 via the submission platform.
Tags: SHROOM-CAP, hallucination detection, LLMs, NLP, IJCNLP-AACL 2025, CHOMPS-2025, scientific content