Research Engineer, Multimodal Reinforcement Learning

DeepMind(13 days ago)

HybridFull TimeSenior$147,305 - $198,720 (estimated)Research

About this role

Research Engineer role within Google DeepMind’s Frontier AI Unit focused on advancing multimodal, image-grounded conversational learning. The position centers on extending meta-training via reinforcement learning to build scalable, semi-verifiable multimodal environments that improve the next generation of models such as Gemini. The role sits at the intersection of research and engineering, bridging core conversational research with visual grounding and infrastructure.

View Original Listing

Required Skills

Reinforcement Learning
Multimodality
Vision-Language
Experimentation
Python
JAX
PyTorch
TensorFlow
Retrieval
Embeddings

+3 more

Qualifications

PhD in Computer Science, AI, or Related Field

About DeepMind

deepmind.google

Google DeepMind is an AI research lab within Alphabet focused on solving intelligence and building safe, general-purpose artificial intelligence to advance science and benefit humanity. It develops cutting-edge machine learning methods—especially deep learning, reinforcement learning and neuroscience-inspired models—and has produced landmark systems such as AlphaGo, AlphaZero and AlphaFold. DeepMind partners with academia, healthcare providers and other Google teams to apply AI to scientific discovery, medicine, energy efficiency and real-world problems. The organization emphasizes safety, interpretability and responsible deployment alongside publishing research and tools for the wider community.