Generative AI
A model that can generate plausible text, images, or molecules has learned something about the structure of its domain. Our generative AI research asks what, exactly, it has learned — and how that knowledge can be made controllable, efficient, and reliable.
- Foundation models
- Large language models
- Diffusion models
- Multimodal generation
- Generative representations
- Model efficiency
- Generative system reliability
Why it matters
Generative models are increasingly used as general interfaces to computation: they write code, summarize evidence, design candidate structures, and act as components inside larger systems. Their usefulness depends on properties that are not captured by fluency alone — factual grounding, consistency, controllability, and an honest account of their own uncertainty.
They are also scientific objects in their own right. A model trained to generate data implicitly represents a probability distribution over that data. Understanding the shape of that distribution tells us both what these systems know and where their knowledge runs out.
Key research questions
How do generative models represent compositional structure, and why does composition fail when it does?
Can we detect, at generation time, when a model's output is unsupported by what it has learned?
What are the trade-offs between autoregressive, diffusion, and hybrid approaches across modalities?
How can generation be steered toward constraints — physical, logical, or stylistic — without retraining?
Which parts of a large model's capacity are necessary, and what can be removed without loss?
Primary subfields
- Foundation models
- Large models pretrained on broad data and adapted to many downstream tasks.
- Diffusion and flow models
- Generating data by learning to reverse a gradual corruption process.
- Multimodal generation
- Joint models of language, images, audio, and structured data.
- Controllable generation
- Conditioning, guidance, and constraint satisfaction during sampling.
- Efficiency
- Distillation, quantization, sparsity, and faster sampling.
Representative methodologies
Likelihood and sample-based evaluation
Combining distributional measures with targeted human and automated review.
Mechanistic analysis
Tracing how specific outputs arise from internal computation.
Constrained decoding
Imposing verifiable structure on generation at inference time.
Compression studies
Measuring which capabilities survive aggressive reduction in model size.
Relationships to other domains
- Natural Language Processing
Language models are the most mature class of generative system.
- Computer Vision
Image and video generation share representations with visual understanding.
- AI Safety
Reliability and evaluation of generative outputs is a central safety concern.
Working on questions in Generative AI?
We welcome conversations with researchers and organizations working on related questions — whether to compare notes, discuss a collaboration, or share work in progress.