Protocol boundary
These protocols are drafts for future validation. They define what should be measured before any empirical claim is made.
Human learning experiment
Compare two groups: one with an assimilation interval after a task and one without it. Keep task type, timing, environment and instructions constant. Measure retention, error rate and quality after a delay.
Cognitive load experiment
Compare completed cycles and interrupted cycles using NASA-TLX, pulse, HRV where ethically collected, and subjective fatigue.
AI simulation
Compare agents with reflection and memory-update phases against baseline agents. Measure stability, repeated mistakes and task quality.
Context for experimental testing
English home · Russian home · View the AI content guide (llms.txt) · View the complete AI context (llms-full.txt)