All deadlines are 11.59 pm UTC -12h (“Anywhere on Earth”).
Large language models (LLMs) are on the rapid rise to empower human researchers in science production at all stages, from the initial conception of a research problem to reporting scientific discoveries (Lu et al., 2024). Recently, the American publishing company Wiley surveyed 5,000 researchers across 70 countries, with the majority supporting LLM adoption in scientific production, although its current usage is limited to writing assistance1. A very recent survey (Eger et al., 2025) by the workshop organization team provides an overview of LLM based science production for users looking to incorporate LLMs into their research workflows, while informing researchers of the current status of this emerging topic, regarding datasets, methods, limitations and future directions.
Recently, there have been several relevant workshops, including NLP4Science, which focuses on LLMs for scientific discoveries in non-CS domains; Intelligent and Interactive Writing Assistants, which focuses on LLMs as writing assistants, and more recently, AI & Scientific Discovery, which addresses LLMs as autonomous agents for accelerating science production across idea generation, experimentation, scientific writing, etc. Our workshop looks for work that (i) integrates human oversight into LLM-assisted research workflows to ensure scientific rigor and ethical standards and (ii) evaluates LLMs for science production and LLM-assisted papers, both of which have not been emphasized in previous workshops. See call for papers for more details. Reference papers are available here.
Email: SciProdLLM@groups.io