Currently, artificial intelligence is undergoing a crucial transformation from "extensive development" to "intensive development", and statistics is the core driving force behind this transformation.

The strategic layout of our department in this direction aims to address the three core challenges faced in the era of large models: the efficiency of model data and learning algorithms, the uncertainty and interpretability of artificial intelligence, and the generalization and universality of artificial intelligence.

Relying on high - level platforms such as the "Interdisciplinary Institute of Data Science", we have gathered research forces with profound accumulations in fields such as deep learning theory, high - dimensional statistics, optimization algorithms, large language models, and artificial intelligence. We are committed to:

Revealing the statistical mechanisms behind the construction of large models, providing theoretical guidance for model design, data usage, and optimization algorithms;

Developing interpretable learning frameworks with theoretical guarantees to promote the trustworthy application of AI in key fields;

Exploring new paradigms for the deep integration of statistical inference and neural networks, reshaping the underlying evolutionary logic of large models towards general intelligence.

Our goal is not merely to pursue "bigger" and "faster" models, but to deeply cultivate "more efficient", "more reliable", and "more universal" models - ensuring that every intelligent decision is effective, substantial, and far - reaching.