PhD Research Scientist Intern - Reinforcement Learning for Diffusion Modelling

📋 Description

• Work with Canva’s AI team on a live, industry-scale project.
• Collaborate with researchers and engineers toward production.
• Research and develop per-layer VLM evaluation and rewards.
• Contribute to layered-generation roadmap and publication efforts.
• Experience hands-on with real data, infra, and deadlines.
• Flexible hybrid work with in-person and remote collaboration.

🎯 Requirements

• PhD in progress, ideally in 3rd year or later.
• Diffusion/flow-matching background with policy-gradient RL.
• Experience fine-tuning VLMs (LoRA) and evaluation prompts.
• Reward modelling, preference optimisation, pseudo-labelling.
• Ability to read and reproduce recent papers quickly.
• Strong written and verbal communication of technical work.

🎁 Benefits

• Internship starting September; 16 weeks, London hybrid.
• Work on products used by hundreds of millions of people.
• Mentors and research-to-production exposure.
• Publication opportunities and potential for patentable work.