Global artificial intelligence (AI) search Start - Up Liner said on the 23rd that three papers completed using its in-house research AI agent were finally accepted to "Agents4Science 2025," the world's first AI author-centered academic event. Among them, the paper by Choi Jae-young, a first-year student at Korea Science Academy of KAIST, was selected as a "Spotlight" paper, one of the top 11.
"Agents4Science 2025" is an international conference hosted by Stanford University in the United States and is the first event where AI participated as the first author of scientific papers and as a peer reviewer. A total of 253 papers were submitted from around the world, and only 48 were finally accepted after an initial AI review and expert evaluation.
Liner supported participants in writing papers in collaboration with AI through the "Liner Max" scholarship program in Aug. As a result, a total of three papers were accepted: ▲ "Real-time thermodynamic feasibility verification of stochastic biochemical models" by Choi Jae-young (selected for Spotlight) ▲ "A study on efficient compute resource allocation strategies for large language models" by ML inference engineer Jeong Beom-jun ▲ "A study verifying the cognitive mechanisms by which AI learns and reproduces social bias" by product head Yoo Yoon-bong.
The research team used research AI tools such as the "Hypothesis Generator," "Citation Recommender," and "Peer Review" to improve the completeness of the papers. They also strengthened the rigor of the study by securing data from about 300 cases using the "Survey Simulator."
Liner assessed this achievement as a case that proves AI's role as an "intellectual colleague" that complements human thinking and creativity. In particular, the fact that a high school student without research experience completed a paper at the level of an international academic conference in just one month showed how AI lowers barriers to entry for research.
Liner CEO Kim Jin-woo said, "The acceptance of three papers, including Choi Jae-young's Spotlight paper, shows that 'Vibe Research' has moved beyond a conceptual attempt to become an empirical innovation," and added, "By having AI take on repetitive and time-consuming tasks, we will create an environment where researchers can focus on deeper inquiry."