Applied Scientist Intern - Monetization Technology - Global Frontier Tech Recruitment Program - 2027 Start (PhD)

t-generation monetization platforms to help millions of customers grow their businesses, utilizing our products like TikTok. Our team develops a wide variety of advertisements for numerous uses including feeds, live streaming, branding, measurement, targeting, search, vertical solutions, creative solutions, and business integrity.

Topic Content: This topic dives deep into TikTok's core global advertising scenarios, driving innovation and implementation of the cutting-edge generative technologies in search, recommendation, and advertising. By deeply integrating foundation models with the advertising business, we address key technical challenges in Large Recommender Models and Large Language Models (LLMs) to build a next-generation intelligent advertising engine with autonomous decision-making capabilities.

Our research covers cutting-edge directions, including Large Recommender Model scaling laws, end-to-end unified modeling, generative full-link technologies (retrieval, ranking, AIGC material generation, bidding), intelligent advertising placement agents, ultra-long sequence modeling, and causal inference. We tackle extreme challenges of trillion-level features and millisecond responses, advancing advertising recommendation toward the foundation model paradigm to achieve dual improvements in monetization efficiency and user experience.

Responsibilities:

1. Explore scaling laws for foundation models in recommendation and advertising, and build a foundation model based on unified multimodal semantic modeling.
2. Build an intelligent ad placement system optimized for users' Long-Term Value (LTV) and long-term ROAS, achieving an optimal balance between commercial value and user experience.
3. Optimize the full-process training and online inference framework for foundation models, balance computing power costs and real-time response performance, and resolve the performance-latency trade-off in real-world deployment.

Qualifications

Minimum Qualifications:

1. Currently pursuing a PhD in Computer Science, Computer Engineering, or a related technical discipline.
2. Modeling experience in one or more of the areas: Ads, Search engine, Recommender System, NLP/CV.
3. Have a solid foundation in algorithms related to LLMs, including but not limited to comprehensive learning and
   practical experience in areas such as single-modal LLM application and deployment.

Preferred Qualifications:

1. Priority will be given to candidates with research results and extensive practical relevant fields, such as outstanding performance in natural language processing, computer vision, data modeling, or algorithm optimization, etc.
2. Excellent programming abilities with a strong command of data structures and fundamental algorithms. For traditional coding roles, proficiency in C/C++ is required; for intelligent coding roles, proficiency in Python is required.
3. Strong publications record in top conferences (e.g., ICLR, NeurIPS, ICML, ACL, EMNLP, NACCL, CVPR, ICCV, and ECCV) is a plus.