Job Title: AI Researcher
Location: Cambridge or London, UK
This is a permanent position with candidates required to do hybrid working in either Cambridge or London.
Our client are looking for AI Researchers specialising in Reinforcement Learning with Human Feedback (RLHF) and Generative AI. In this role, you will design and optimise the algorithms that align large-scale generative models with human preferences, ensuring they are safe, controllable, and capable of producing high-quality outputs across multiple modalities. You’ll sit at the intersection of RL, LLMs, and generative modelling, helping us build the next generation of foundation models
Responsibilities:
Develop and refine RLHF algorithms for large language and generative models.
Research and implement deep reinforcement learning methods (policy gradients, actor-critic, off-policy learning) for model alignment.
Train, fine-tune, and evaluate LLMs and diffusion models at scale.
Design experiments to align generative outputs with human and organisational preferences.
Collaborate with researchers, engineers, and human feedback teams to build scalable alignment pipelines.
Publish findings in top-tier AI conferences and contribute to open-source frameworks.
Key Requirements:
PhD in Computer Science, Machine Learning, or related field.
Publications at NeurIPS, ICML, ICLR, ACL, or related venues.
Deep expertise in Reinforcement Learning (policy optimisation, reward modelling, RLHF).
Hands-on experience training/fine-tuning generative models (LLMs, diffusion, transformers, GANs).
Strong knowledge of deep learning frameworks (PyTorch, JAX, TensorFlow).
Proficiency in Python and standard ML libraries.
Solid foundations in probability, optimisation, and statistics.
Experience working with large-scale distributed training on GPUs/TPUs.
If this sounds of interest, please reach out to
