About the job
The algorithm team is responsible for developing state-of-the-art computer vision, NLP and multimodality models and algorithms to protect our platform and users from the content and behaviors that violate community guidelines and related regulations. With the continuous efforts from our team, TikTok is able to provide the best user experience and bring joy to everyone in the world.
Responsibilities
1. Leverage multimodal large models to explore few-shot and zero-shot strategies for content safety scenarios, and build moderation models with strong generalization capabilities.
2. Participate in reinforcement learning–based data mining, and help design Chain-of-Thought (CoT) annotation frameworks to improve the model’s understanding of complex risks.
3. Build risk ranking and recall systems to enhance coverage and accuracy in identifying high-risk content.
4. Collaborate with product and policy teams to drive real-world deployment and performance optimization of moderation algorithms.
Qualifications
Minimum
1. Currently pursuing a Master degree with a background in computer science, machine learning, or similar fields;
2. Solid foundation in machine learning and deep learning; familiarity with multimodal modeling; hands-on experience with large model projects is a plus.
3. Interest in content safety, with an understanding of the challenges in risk identification within moderation workflows.
4. Proficient in end-to-end algorithm development, including data processing, modeling, and evaluation; experience with PyTorch or TensorFlow is required.
5. Able to commit to working for 12 weeks during the Summer of 2026.
Preferred
Strong learning ability, clear communication, and a strong sense of teamwork and responsibility.