Conference on Cross-Disciplinary Studies and Large Language Models
Exploring how humans and AI co-create the future of knowledge
CROSS-LLM 2025 is the first interdisciplinary conference designed to explore how Large Language Models (LLMs) can contribute to—and challenge—human research across domains.
The conference experiments with hybrid authorship: human-written research papers are paired with AI-generated counterparts on similar topics. Both are then evaluated by domain experts under a randomized, double-blind setup.
Recently published interdisciplinary papers from social science, natural science, and computational fields are selected to represent state-of-the-art human scholarship.
Large Language Models generate parallel research ideas or drafts inspired by the same topics and datasets. These serve as computational thought experiments for interdisciplinary exploration.
All submissions—human and AI—are randomly distributed to expert reviewers across disciplines. Reviewers evaluate novelty, rigor, and insight without knowing whether a paper was authored by a human or an AI.
Submissions include short papers (up to 4 pages) or extended abstracts (2 pages). Both human and AI authorships are welcome. Each submission must include:
Reviews are double-blind and randomized. Domain experts from computer science, social sciences, humanities, and natural sciences assess each paper for:
Email: info@cross-llm.org