Research Topics

Robust and Trustworthy AI

Robust and Trustworthy AI refers to the development of artificial intelligence systems that can reliably operate under various conditions, including unexpected or adversarial environments, while ensuring ethical, transparent, and fair outcomes. Robust AI focuses on creating models that are resilient to errors, biases, and uncertainties, ensuring consistent performance. Trustworthy AI, on the other hand, emphasizes the importance of AI systems that users and society can trust, which involves aspects like explainability, fairness, accountability, and alignment with human values.

Pre-trained Models (e.g., LLMs)

Pre-trained Models refer to AI models that are trained on large datasets prior to being fine-tuned for specific tasks. These models, such as LLMs (Large Language Models) and VLMs (Vision-Language Models), are increasingly used to leverage their general understanding in various domains.

In the case of LLMs, they are widely employed for natural language processing tasks, including text generation, summarization, and question answering. VLMs, on the other hand, integrate both visual and textual information, making them valuable for tasks like image captioning, visual question answering, and multimodal reasoning.

Research on utilizing these models focuses on how to efficiently adapt and fine-tune them for specialized tasks while maintaining their robustness and efficiency. Techniques such as transfer learning, prompt engineering, and multimodal fusion are key areas of study to enhance their performance across diverse applications, from language understanding to image-text interaction.


AI for X (e.g., AI for energy)

AI Application refers to the practical use of artificial intelligence technologies to solve real-world problems across various industries. These applications span a wide range of domains, including healthcare, where AI assists in diagnostics and personalized medicine; finance, where it is used for fraud detection and algorithmic trading; and energy, where AI optimizes resource management and efficiency. AI applications also extend to autonomous systems, such as self-driving cars, and creative industries, where AI generates art, music, and written content. By automating tasks, improving decision-making, and uncovering new insights, AI applications are transforming industries and daily life.