Published
Book One · From Document to Evidence
How a document becomes disciplined evidence—and why interpretation must wait.
Our Work
Adventures in AI
Five connected investigations into how artificial intelligence learns, reasons, and represents knowledge.
4 published · 1 in progress
Published
Published
Book Two · From Evidence to Interpretation
Teaching, persistence, and the difference between changed behavior and reusable knowledge.
Published
Book Three · From Interpretation to Measurement
Practical ways to measure what changes inside a model when it learns—including updates in model weights.
Published
Book Four · From Measurement to Knowledge
Controlled evaluations of how an AI uses acquired knowledge in unfamiliar situations—without claiming to prove how knowledge is stored inside the network.
Work in Progress
Book Five · From Knowledge to Representation
Begins investigating how knowledge may be represented inside neural networks. Introduction and Chapter One are complete; later chapters are not yet written. The full book is not for sale.
Free research tool
AI Model Inspector
A free research tool for examining supported AI model checkpoints and published measurements.
For educators
Adventures in AI presents experimental case studies for courses and seminars in AI, machine learning, model evaluation, and reasoning. Educational organizations may request a complimentary instructor review copy.
Support the work
Free previews are available for every volume. Purchases of Books One–Four support LCC Research Lab research and publication. Book Five remains a free Reader Preview while the volume is in progress.