Description

A project demonstrating 'loop engineering', a system that moves beyond simple prompting to use evaluators, memory, and stop rules for iterative refinement. It includes a simulator ('toy sim') that analyzes how AI agents report gains versus real performance, accompanied by a code runner and technical field guide.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.