Considering the rapid development of LLM technologies and the great importance and challenge of LLM’s safety and evaluation, Journal of Computer Science and Technology (JCST) and NLPCC 2025 jointly call for papers on safety and evaluation of LLMs. Selected papers will be published as a special section of a JCST 2025 regular issue, and the papers will be invited to be presented (either orally or as poster) in the main conference of NLPCC 2025, which will be held on August 8-10, 2025, in Urumqi, China.

The relevant topics include (but are not limited to):

● Hallucination and misinformation

● Toxicity, fairness, bias, and ethics

● Red-teaming and jailbreaking

● Watermark and privacy for LLMs

● Adversarial attacks and defenses

● Safety alignment

● LLM/NLG evaluation benchmarks and metrics

● Human evaluation

● Applications in sensitive domains (e.g., medical, legal, finance)

● Xiaojun Wan, Peking University, wxj@pku.edu.cn

● Derek F. Wong, University of Macau, derekfw@um.edu.mo

● Yue Zhang, Westlake University, yue.zhang@wias.org.cn

● Submission deadline: 30 April, 2025

● First round of reviews: 30 May, 2025

● Revision deadline: 15 June, 2025

● Notification of acceptance/rejection: 30 June, 2025

● Final manuscript due: 15 July, 2025

● Paper Presentation in NLPCC 2025: 8-10 August, 2025

● Expected Publication date: Sept. 2025

● Submissions must use JCST submit template (https://jcst.ict.ac.cn/Download_Templates), and should be 8 to 12 pages in length (excluding the appendix).

● The submitted papers must not be extended versions of previously published papers, and must be original submissions, not under review in other venues.

● All manuscripts will be submitted through JCST submission system https://mc03.manuscriptcentral.com/jcst, selecting the manuscript type “Special Section on Safety and Evaluation of LLMs”.

● All submissions should be written in English and will undergo a double-blind peer review process.