Emergent Intelligent Lab
EIL is an independent ML research lab advancing core algorithms and LLM behavioral research using first-principles approaches to architecture design and diagnostic systems — for building robust AI systems.
Research
First-principles work on both sides — to architecture design on one, and to the diagnostic systems we use to study large language models on the other.
First-principles architecture design — cross networks, mixture-of-experts, low-rank parameterizations, and the training infrastructure underneath them.
See projects
Diagnostic systems for large language models — multi-agent experiments that treat LLMs as participants and surface biases, capabilities, and failure modes.
See projects
Projects
Active projects across both pillars. Each card links to a deeper write-up.
Cross-network architectures with low-rank decomposition and mixture-of-experts gating — explicit feature-interaction layers for deep models.
Our shared ML platform: Metaflow flows, AWS Batch compute, MLflow tracking, and reproducible PyPI dependency management across local and cloud runs.
Behavioral interference experiments measuring egocentric bias, naïve realism, and cognitive inflexibility across frontier LLMs — with a procedural prompt debiasing library adapted from social-cognition research.
A reusable scaffold of four specialized LLM agents — Lead Contributor, Positioning, Method, and Experiment owners — that collaborate to draft, review, and assemble a research paper end-to-end.
Contact
Interested in our work, want to collaborate, or curious about a specific project? Drop us a line.
[email protected]
General inquiries, collaborations, and press.
Code
Our open repositories
Each project links out to its repository. New work lands there first.
A note: EIL is a small, independent research group. We're selective about collaborations but try to read every well-targeted message — please include context about which pillar your interest connects to.