The field of artificial intelligence research is moving towards the development of more dynamic and interactive systems for automated scientific discovery. Recent advancements have focused on creating multiagent frameworks that can adapt to intermediate findings and incorporate human feedback, enabling the transformation of automated research into continual research programs. These systems have the potential to facilitate broader adoption of automated research across scientific domains, allowing practitioners to deploy interactive multiagent systems that autonomously conduct end-to-end research. Noteworthy papers include:

• A multiagent framework that features fully dynamic workflows and a modular architecture, enabling seamless customization and comprehensive infrastructure for automated research.
• An autonomous pipeline that converts scientific papers into interactive and multimedia-rich academic homepages, outperforming end-to-end baselines and achieving the Pareto-front in academic webpage generation.
• A structured framework for agentic system composition that enables a composer agent to systematically identify, select, and assemble an optimal set of agentic components, streamlining the assembly of agentic systems and facilitating scalable reuse of resources.
• A novel multi-agent system that collaboratively crafts project webpages from research papers, generating high-quality and visually appealing pages with remarkable efficiency.