Liner AI: World's First AI Author with 3 Conference Papers Accepted

Global AI search startup Liner has achieved a remarkable feat, demonstrating the innovative capabilities of its research AI agents by having a total of three papers accepted at 'Agents4Science 2025', an AI-author-centric academic event. In particular, the paper by Choi Jae-young, a first-year student at Korea Science Academy of KAIST, made a splash by being named a 'Spotlight' paper, the highest honor, selected from only the top 11 papers worldwide. 'Agents4Science 2025', hosted by Stanford University in the U.S., was the world's first experimental academic event where AI participated as a first author and peer reviewer of scientific papers, serving as a pioneering platform to explore AI's contribution to research. Out of 253 papers submitted from around the world, 80 were selected after initial screening by large language models (LLMs), and only 48 were finally accepted through a rigorous qualitative evaluation by expert reviewers, making it a prestigious competition. The event concluded successfully on the 22nd, U.S. local time, with the presentation of the Spotlight papers. Last August, Liner actively supported future researchers to write papers in collaboration with AI and participate in this international academic conference by operating a scholarship program that provided free access to its self-developed 'Liner Max'. As a result of this support program, an impressive achievement was made with three papers being finally accepted. The accepted papers are Choi Jae-young's ‘Real-time Thermodynamic Validation of Stochastic Biochemical Models’, Jung Bum-jun (Liner ML Inference Engineer)'s ‘A Study on Efficient Computational Resource Allocation Strategies for Large Language Models’, and Yoo Yoon-bong (Liner Head of Product)'s ‘A Study on Cognitive Mechanisms for AI to Learn and Reproduce Social Biases’. These three authors actively utilized Liner's research AI agents throughout the entire paper writing process. They refined their research topics using the 'Hypothesis Generator' and quickly secured extensive reference materials with the 'Citation Recommender'. Furthermore, they significantly enhanced the completeness of their papers by undergoing multiple review processes in a short time using the 'Peer Review' agent. In particular, Head of Product Yoo Yoon-bong efficiently collected over 300 pieces of data through the 'Survey Simulator', adding depth to the research. Choi Jae-young stated, "The most meaningful part is that thanks to Liner's AI agents, I was able to complete the paper myself without anyone else's help and challenge myself at a global academic conference." He added, "Ultimately, the output of AI varies depending on the thinking ability of the person asking the questions, and now, the ability to pose deep and fundamental questions has become a core competency in research." Kim Jin-woo, CEO of Liner, emphasized, "The acceptance of a total of three papers, including Choi Jae-young's Spotlight paper, demonstrates that 'Vibe Research' pursued by Liner has firmly established itself as empirical innovation beyond a mere concept." He further presented a vision for the future of AI-driven research, stating, "By having AI handle repetitive and time-consuming tasks, researchers can focus on deeper, essential thinking and creative inquiry, thus ushering in a true era of research."