Large Signal Model
From an observation to a predictive state
An exact archive preserves an observation. A useful predictive state must also retain what the declared future questions need. Start with that distinction, then follow the construction and its evidence.
Read the mathematical body · Find a proposition or proof
Explore the visual atlas · 17 plates, original diagrams, and an exact closure explorer.
A first reading path
- Distinguish observations, archives, and predictive states
- See when a summary supports a reduced transition
- Follow the complete small signal example
- Identify a recurrence from observed outputs
Study the full construction
Continue through learning and stability, controlled stochastic realization, certified probability domains, continuous observations, and robustness under data uncertainty. The contents group the 100 sections into eight parts.
Open all chapters · Try the Signal laboratory
Inspect the recorded formal scope
The retained coverage map records complete mappings for 87 registered proposition cores under explicit assumptions. Each HTML proposition links to its written proof and its recorded scope. The publication conversion adds no compiler run or mathematical review.
Read the scope of each proposition
- Inspect the mathematical Lean sources
- Read the proposition coverage map
- Read the Lean companion guide
- Inspect the accepted verification summary
- Full accepted execution record on GitHub
Explore the software and reference examples
The PR-LASM software package has its own implementation and assessment record. Use its documentation for setup and runtime details; the mathematical coverage count does not establish empirical model performance.
Open the PR-LASM software work page
Run the retained reference examples
Original publication and atlas
The 174-page PDF retains the full publication layout, additional glossary, and formal registers. The visual atlas is available in HTML alongside its original 19-page PDF.
Open the visual atlas · Open publication and companion links