Senior Researcher, Foundations of Generative AI: Microsoft Research (AI Frontiers Lab)

The AI Frontiers lab focuses on expanding the Pareto frontier of AI capabilities, efficiency, and safety. This role involves leading efforts to discover and build the foundations of generative AI through new representations and objectives. Research areas include small language models (Phi, Orca), new architectures (transformers, state space models), agentic AI systems, and multimodal models (VLM, VLA).

Specific Research Focus Areas

New architectures and learning methodologies for proactive agents.

Test-time training for infinite context windows.

Active visual reasoning for faster image processing.

World models of user/system interactions for imagination search.

Multi-scale temporal reasoning and planning.

Continual learning and adaptation at human speed.

Responsibilities

Develop, prototype, and evaluate cutting-edge research ideas.

Rapidly test new architectures and publish results in top-tier venues.

Collaborate with engineers and product groups to deliver solutions for real-world problems.

Qualifications Minimum Requirements:

Doctorate (or currently pursuing) in Computer Science or a relevant field, or equivalent experience.

Preferred Requirements:

Doctorate plus 2+ years of related research experience.

A research program demonstrated by public artifacts (models, code) or publications at conferences like NeurIPS, ICML, ICLR, ACL, CVPR, etc.

2+ years of experience with Python, PyTorch, TensorFlow, or HuggingFace.

Deep understanding of frontier architectures, specifically transformers and state space models.

Hands-on experience with LLMs, VLMs, or VLAs, including pre-training and fine-tuning.

2+ years of experience building and optimizing large-scale ML training pipelines.

Compensation

New York City Area Range: $158,400 – $258,000 per year.

Standard U.S. Range: $119,800 – $234,700 per year.

Pay is subject to location and experience and may include other compensation or benefits.