Visual Generation & Multimodal Evaluation Researcher Graduate (AML-Ark-US) - 2027 Start (PhD)

Founded in 2012, ByteDance's mission is to inspire creativity and enrich life. With a suite of products including TikTok, Lemon8, CapCut, and Pico, ByteDance makes it easier and more fun for people to connect with, consume, and create content.

About the role

The Applied Machine Learning Ark team combines system engineering and machine learning to develop and operate Large Language Model (LLM) service platforms that offer businesses Model-as-a-Service (MaaS) solutions, serving both large model providers and downstream users. The US team drives the design, development, and operation of MaaS solutions across the US and international markets outside mainland China.

We are building full-stack, end-to-end solutions spanning text and multimodal LLM algorithms, LLM training/fine-tuning/inference frameworks, prompt engineering, model alignment, and intelligent agent systems. Beyond model serving, we operate large-scale log analytics pipelines that process massive volumes of invocation logs from text models, multimodal models, and agent systems—extracting usage patterns, quality signals, and actionable insights to inform model improvement, system optimization, and product decisions through continuous, data-driven feedback loops.

Responsibilities

• Build evaluation systems for image and video models/agents, covering generation quality, instruction following, multimodal understanding, and safety.
• Develop automated metrics and model-based evaluators, and design reproducible human evaluation protocols.
• Design and develop video generation/debugging agents that orchestrate multi-step creative workflows.
• Build large-scale image and video data pipelines, and turn evaluation findings into model and product improvements.

Requirements

• Individuals who are completing or have recently completed a PhD's degree in Computer Science, Artificial Intelligence, Machine Learning, Computer Vision, or a related field.
• Solid foundation in deep learning and computer vision, including generative modeling fundamentals.
• Practical experience in at least one of: visual generation, multimodal LLMs, video understanding, or visual quality assessment.
• Strong Python skills and proficiency with PyTorch or an equivalent framework, or multimodal evaluation framework.
• Demonstrated research or engineering ability through publications, substantial projects, internships, or open-source work.

Preferred Qualifications

• Publications at top-tier vision or ML venues, e.g., NeurIPS, ICML, CVPR, ICCV, ECCV, etc.
• Hands-on experience with modern visual generation stacks, including diffusion-based models and their post-training.
• Familiarity with visual generation benchmarks, or experience building evaluation frameworks.
• Experience applying agent frameworks to creative workflows, or working with large-scale video data infrastructure.

Pay

The base salary range for this position is $153,900 - $300,960 annually. Compensation may vary depending on qualifications, skills, competencies, experience, and location. Base pay is one part of the total package, which may include additional discretionary bonuses/incentives and restricted stock units.

Benefits

• Day one access to medical, dental, and vision insurance.
• 401(k) savings plan with company match.
• Paid parental leave, short-term and long-term disability coverage, and life insurance.
• Wellbeing benefits.
• 10 paid holidays per year.
• 10 paid sick days per year.
• 17 days of Paid Personal Time (prorated upon hire with increasing accruals by tenure).