Concept Based Model

Concept-based models aim to enhance the interpretability and trustworthiness of deep learning models by incorporating human-understandable concepts into their architecture and decision-making processes. Current research focuses on developing novel architectures, such as hybrid and relational concept-based models, and algorithms that leverage disentangled representations, large language models, and self-supervised learning to efficiently learn and utilize these concepts, even with limited data or human annotation. This work is significant because it addresses the "black box" nature of many deep learning systems, improving their transparency, reliability, and ultimately, their applicability in high-stakes domains requiring explainable AI.