Student Researcher [Seed LLM Horizon - Multi-turn Tool Use] - 2026 Start (PhD)

Job Description:

About the team The Seed LLM Horizon Team is dedicated to cutting-edge research, driven by a mission to push the boundaries of model intelligence, and fuelled by a long-term vision and unwavering commitment. The team is dedicated to developing the next-generation agent foundation model and building self-evolving、personalized Agent. We are seeking passionate and self-driven researchers who share our vision to collaborate on agent research. PhD internships at ByteDance provide students with the opportunity to actively contribute to our products and research, and to the organization's future plans and emerging technologies. Our dynamic internship experience blends hands-on learning, enriching community-building and development events, and collaboration with industry experts. Applications will be reviewed on a rolling basis - we encourage you to apply early. Please state your availability clearly in your resume (Start date, End date). Responsibilities - Enable models to perform deep usage of professional tools (e.g., search, code-interpreter) to solve complex problems. - Develop approaches to generalize model abilities to millions of out-of-distribution (OOD) tools and scenarios. - Scale up multi-turn tool-use training tasks and explore effective training methods. - Address challenges of long-horizon, multi-turn tasks in reinforcement learning

Candidate Requirements:

Minimum Qualifications: - Currently pursuing a PhD in Computer Science, Software Engineering, Machine Learning, or a related field. - Research experience in one or more of the following: reinforcement learning, LLM agents, memory systems, tool use, or interactive learning. - Strong coding skills and proficiency with modern deep learning frameworks. - Demonstrated ability to conduct independent research, with publications in top-tier ML/AI conferences such as NeurIPS, ICML, ICLR, ACL, EMNLP etc. Preferred Qualifications: - Experience with long-horizon reasoning, multi-turn tasks, or asynchronous agent behavior. - Familiarity with agent evaluation, personalization, or real-world tool integration. - Background in building or analyzing large-scale agent training pipelines. - Ability to collaborate effectively in a fast-paced, research-driven team environment