The 35th International Joint Conference on Artificial Intelligence (IJCAI), held jointly with the European Conference on Artificial Intelligence (ECAI), took place in Bremen, Germany, from August 15 to 21, 2026.

Founded in 1969, IJCAI is one of the world’s longest-running and most prestigious comprehensive conferences in artificial intelligence. It brings together AI researchers from around the world and covers major areas including machine learning, intelligent agents, knowledge representation and reasoning, computer vision, natural language processing, and robotics. The conference serves as an important international forum for presenting cutting-edge research and emerging trends in artificial intelligence.

The IJCAI Early Career Spotlights (ECS) program recognizes promising early-career researchers with outstanding contributions across major areas of artificial intelligence. Selected researchers are invited to present their research directions, representative achievements, and academic visions in a dedicated conference session. Since its establishment in 2016, the IJCAI Early Career Spotlights program has featured scholars including Yejin Choi and Stefano Ermon from Stanford University, Ruslan Salakhutdinov and J. Zico Kolter from Carnegie Mellon University, Bo An from Nanyang Technological University, and Jun Zhu and Zhiyuan Liu from Tsinghua University.

At IJCAI-ECAI 2026, Yaqing Wang, Associate Professor at the Beijing Institute of Mathematical Sciences and Applications (BIMSA), was selected for the Early Career Spotlights program and invited to deliver a spotlight talk. Fourteen early-career researchers worldwide were selected for this year’s program, and Yaqing Wang was the only selected researcher based in mainland China.

Yaqing Wang’s research focuses on machine learning and artificial intelligence. She investigates how AI systems can learn efficiently and adapt rapidly under limited data, limited supervision, and dynamic environments, with the aim of advancing scientific discovery and real-world industrial applications. She received her Ph.D. in Computer Science and Engineering from the Hong Kong University of Science and Technology in 2019 and previously worked as a Senior Researcher at Baidu Research.

Yaqing Wang has published 40 papers in leading international machine learning and artificial intelligence conferences and journals, including ICML, NeurIPS, ICLR, AAAI, IJCAI, TPAMI, JMLR, and TIP. Her work has received more than 6,000 citations. She was selected for the Beijing Nova Program in 2025, the AAAI New Faculty Highlights program in 2026, and the IJCAI Early Career Spotlights program in 2026. She was also named to the World’s Top 2% Scientists list for two consecutive years, in 2024 and 2025. She currently serves as Vice Chair of the IEEE Computational Intelligence Society Task Force on Data-Efficient Agentic Learning, Executive Editor of Neural Networks, and an Editorial Board Member of Machine Learning. She also serves as an Area Chair for ICLR and ACL Rolling Review and as a Senior Program Committee Member for AAAI.

In her spotlight talk, titled “Toward Data-Efficient Intelligence: From Few-Shot Learning to Agentic Systems,” Yaqing Wang presented a systematic account of her long-term research on data-efficient artificial intelligence. She summarized the evolution of her research from few-shot learning, meta-learning, and in-context learning to data-efficient agentic learning, highlighting the key results and perspectives developed throughout this work. Drawing on her team’s recent theoretical and methodological studies, Yaqing Wang explained how prior knowledge can enable AI systems to generalize reliably from limited data and discussed the intrinsic connection between in-context learning and meta-learning. She further introduced Data-Efficient Agentic Learning (DEAL), a research framework for investigating how agents can rapidly adapt and make effective decisions with limited supervision and interaction.

As AI evolves from single-step prediction toward multi-step interaction and autonomous decision-making, the notion of data is expanding beyond static training samples to include the experience, feedback, and trajectories generated through interactions between agents and their environments. This shift makes data scarcity an increasingly prominent challenge.Yaqing Wang argued that the effective use of prior knowledge, limited experience, and structural mechanisms is essential for improving the generalization, adaptation, and decision-making capabilities of AI systems. Mathematical theory, meanwhile, can provide deeper insight into the mechanisms underlying learning and adaptation, establishing a theoretical foundation for more data-efficient, reliable, and interpretable AI. These advances may further support applications in scientific computing, biomedicine, and personalized agents.