You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Create a first-class research-notes system with institute cross-linking and trust labels #4
The site has a blog, but the current research program needs a structured publication layer rather than isolated portfolio posts.
Create a research-notes system that can publish ongoing machine-mathematics work with clear evidence status, reproducibility links, and cross-linking to the Mnemosyne Research Institute.
Required content model
Each research note should support:
title, summary, publication date, last-reviewed date
authors and institute affiliation
research program
tags and theorem/problem identifiers
evidence status
repository, commit, issue, dossier, and artifact links
environment/toolchain hashes where relevant
prior-art and citation-review status
disclosure status for proof bodies and benchmark-protected material
related notes and supersession links
Trust labels
Support visible, consistent labels such as:
kernel_verified
certificate_checked
externally_cited_unreviewed
human_reviewed
experimental_evidence
informal_strategy
open_obligation
superseded
These labels must describe evidence without styling unverified claims as equivalent to proof.
Site changes
Add a dedicated /research or /notes index separate from generic projects.
Add program filters for Proof Search, MathCorpus, open conjectures, Erdős campaigns, ANLU, OTCF, and applied verification.
Add RSS/feed support if not already present.
Add social preview metadata and canonical URLs.
Add institute branding and links on every research note.
Add a Cite this note block with a stable citation format.
Add a visible correction/supersession mechanism.
Initial notes to support
Shift-8 forensic proof-search case study
Erdős 647 campaign report
Proof Search MCP architecture and zero-to-one trajectory thesis
MathCorpus machine-curriculum and RL-transition design
Wall mapping as a research objective
Acceptance criteria
Research notes are distinguishable from product/project pages.
Evidence labels are rendered consistently and documented.
A note can be updated without silently changing its original publication date.
Activity