AITRIX's three papers have been accepted simultaneously at ICLR 2026.

AITRICS (CEO Gwangjun Kim), a medical artificial intelligence company, announced on the 5th that three papers from its research team were simultaneously accepted at ICLR 2026, an international academic conference in the field of AI.

ICLR is a leading AI conference, drawing participants from researchers and global IT companies worldwide. It is recognized as a premier academic conference where the latest machine learning research findings are shared. Approximately 19,000 papers were submitted to ICLR 2026, and only the top 28% were accepted after review.

The three papers accepted this time are Delta-XAI, an explainable AI technology that quantitatively explains the causes of changes in patient risk prediction scores; isomorphic self-supervised learning technology that improves performance stability in various medical environments by reflecting data change structures; and class-imbalanced semi-supervised learning research that improves data imbalance problems such as rare diseases through semantic grouping.

These studies are being evaluated for their academic significance and potential for industrial application, as they present methodologies to address key challenges in the field of medical AI: difficulties in interpreting results, performance degradation due to data transformation, and limitations in learning from rare data.

AITRIX plans to present its research results at the ICLR 2026 conference, which will be held in Rio de Janeiro, Brazil in April, and plans to utilize the technologies to enhance the performance of its medical AI products.

AITRIX plans to continue developing AI technology that can be utilized in the medical field by using problems that arise in actual clinical environments as a starting point for research.

• See more related articles