The field of artificial intelligence is witnessing a significant shift towards the development of more efficient and effective information retrieval and reasoning systems. Recent research has focused on creating agents that can autonomously browse the web, synthesize information, and return comprehensive citation-backed answers. This has led to the emergence of new paradigms such as Agentic Deep Research, which integrates autonomous reasoning, iterative retrieval, and information synthesis into a dynamic feedback loop. Notable advancements include the development of reinforcement learning frameworks that enable large language models to perform on-demand, multi-turn search in real-world Internet environments, and the introduction of novel evaluation benchmarks and methodologies that can assess the performance of agentic search systems. Key innovations in this area include the integration of exploration and reasoning, selectively scoped memory mechanisms, and the use of semantic quality-driven prioritization techniques for web crawling. Some noteworthy papers in this area include MEM1, which introduces an end-to-end reinforcement learning framework that enables agents to operate with constant memory across long multi-turn tasks, and Mind2Web 2, which proposes a novel Agent-as-a-Judge framework for evaluating agentic search systems. MMSearch-R1 is also notable for its end-to-end reinforcement learning framework that enables large multimodal models to perform on-demand, multi-turn search in real-world Internet environments.