• Research Mastering Stratego, the classic game of imperfect information
• Research DeepMind’s latest research at NeurIPS 2022
• Research Building interactive agents in video game worlds
• Research How undesired goals can arise with correct rewards
• Science Discovering novel algorithms with AlphaTensor
• Research Building safer dialogue agents
• Research From motor control to embodied intelligence
• Research DeepMind’s latest research at ICML 2022
• Research Human-centred mechanism design with Democratic AI
• Research Evaluating Multimodal Interactive Agents