Skip to content
JournalsWorldThe Global Research Discovery Platform
Featured Dataset

Lean 4 Formalized Proof Artifacts for Local Softmax and Global Weights in Non-Boolean Event Structures

This supplement records the Lean 4 formalization used to verify the finite claims about gluing, generalized-softmax coordinates, and the pentagon half-weight. The project was checked with Lean v4.30.0 and mathlib v4.30

👤
CreatorSvozil, Karl
📅
Published2026-05-31
🔗
DOI10.5281/zenodo.20479152
📊
Downloads2
⚖️
Licensecc-by-4.0
File Size7.7 KB
Data TypeDataset
Published2026
Licensecc-by-4.0
Total Views25
Total Downloads2
This supplement records the Lean 4 formalization used to verify the finite claims about gluing, generalized-softmax coordinates, and the pentagon half-weight. The project was checked with Lean v4.30.0 and mathlib v4.30.0; the command lake build completed successfully, and the source contains no sorry, admit, or axiom.
The formalization separates three layers:
  • First (Core.lean) defines finite event structures, admissible weights, positive coordinates, and the representation theorem for scaled link-function preimages.
  • Second (Gluing.lean) formalizes local context-wise distributions, single-valuedness, the normalizer-ratio calculation, and the theorem that single-valued local distributions glue to a global admissible weight.
  • Third (Pentagon.lean) formalizes the pentagon example: the half-weight is admissible, every two-valued state and every finite classical mixture obeys the cyclic bound 2, the half-weight has cyclic sum 5/2, and 5/2 > √5. The same file also verifies the positive softmax-coordinate path and its boundary endpoint at the half-weight.

📤 Share this page

Found this useful? Share it with your network.

✓ Link copied! Paste it on ResearchGate / Academia.edu
📦
Lean 4 Formalized Proof Artifacts for Local Softmax… (Full Dataset)7.7 KB
⬇
📄
ReadmeVia DOI record
↗

Files are hosted on the source repository. Click download to access the full dataset.

Svozil, Karl (2026). Lean 4 Formalized Proof Artifacts for Local Softmax and Global Weights in Non-Boolean Event Structures. https://doi.org/10.5281/zenodo.20479152