Research
Pushing the boundaries of artificial intelligence through rigorous research and open collaboration.
Focus Areas
Large Language Models
Advancing the capabilities of language models through novel architectures, training techniques, and alignment methods.
Multimodal Learning
Developing AI systems that can understand and generate across multiple modalities including text, images, and audio.
AI Safety & Alignment
Ensuring AI systems behave as intended and remain aligned with human values as they become more capable.
Reinforcement Learning
Creating agents that can learn complex behaviors through interaction with their environment.
Publications
Scaling Laws for Neural Language Models
Nocetic Research Team
Constitutional AI: Harmlessness from AI Feedback
AI Safety Team
Efficient Transformer Architectures for Long-Context Tasks
Architecture Research Group
Multimodal Reasoning in Large Models
Multimodal Team
Open Research
We believe the biggest breakthroughs come from collaboration. That's why we actively publish our research, share our findings, and work with the global AI community.
Our researchers regularly publish at top-tier conferences and collaborate with leading academic institutions worldwide.
Collaborate With Us