GENERATIVE EXPLORATION MODEL (GEM)

To speed up the identification of promising new OLED materials, CYNORA developed the “Generative Exploration Model” (GEM). This AI-based tool adapts an Evolutionary Algorithm approach and modifies molecules iteratively until favorable properties are achieved. The GEM scans the chemical universe and generates tens of thousands of molecules during each run. In concert with powerful and accurate Machine Learning models to estimate molecular properties from molecular structure, the GEM can identify the most promising molecules within minutes.
To optimize the GEM’s capabilities, CYNORA leverages synergies between its in-house synthesis team, computational chemists and device physicists. The combination of technology and expertise enables the GEM to address challenges like synthetic accessibility and device properties already during the in-silico generation process.