Realistic Mathematichs Education in Deep Learning Literature Review: Strengthening Critical Thinking and Problem Solving

Mathematics learning in primary and secondary schools still faces major challenges in the form of low critical thinking skills and student problem-solving due to conventional approaches that are memorized and less contextual. This study aims to comprehensively analyze the integration of Realistic Mathematics Education (RME) in the framework of Deep Learning as an innovative strategy to strengthen students' critical thinking and problem-solving skills. This study uses a systematic literature review method with the PRISMA approach. Data was obtained from 15 scientific journal articles published by 20242026 taken from the Google Scholar, Sinta, Garuda, and accredited national journals databases. The analysis was carried out through narrative synthesis and thematic analysis to identify synergies between the two approaches. The results of the review show that the integration of RME and Deep Learning produces a significant synergistic effect. RME provides real context and local culture as a foundation, while Deep Learning adds adaptivity, real-time feedback, visualization, and deep reflection. This combination consistently improves concept understanding, critical thinking, logical thinking, and problem-solving skills at various levels of education (elementary to high school). The integration of RME in Deep Learning has proven to be an adaptive, contextual, and meaningful learning model, in line with the Independent Curriculum. This hybrid approach is recommended to be widely applied to improve the quality of mathematics education in Indonesia.