NRR-Coupled (cp): Dependency-Consistency Evaluation for Coupled State Updates

NRR-Coupled (cp) is a client-side extension for dependent-candidate state updates under a fixed LLM interface (\sigma/\delta, target).
The repository includes manuscript, specification, simulation code, and reproducibility artifacts.

Part of the Non-Resolution Reasoning (NRR) research program.

NRR Series Hub (Start here)

For the cross-paper map and current series links, start here:

• NRR Series Hub

NRR is not an anti-LLM framework.
NRR does not replace standard LLM use.
NRR optimizes when to commit and when to defer, under explicit conditions. Series numbering policy: paper3 is permanently skipped and never reused.

Quick links

• arXiv: pending (pre-submission; no public URL yet)
• manuscript (current): manuscript/current/paper6-nrr-coupled-v19.tex
• checksum manifest: manuscript/current/checksums_sha256.txt
• specification: spec/nrr-coupled_spec.md
• simulation: repro/coupled_state_sim.py
• reproducibility guide: reproducibility.md

DOI

• 

Repository structure

nrr-coupled/
|-- README.md
|-- LICENSE
|-- requirements.txt
|-- reproducibility.md
|-- manuscript/
|   `-- current/
|       |-- paper6-nrr-coupled-v19.tex
|       |-- paper6-nrr-coupled-v19.pdf
|       `-- checksums_sha256.txt
|-- spec/
|   `-- nrr-coupled_spec.md
`-- repro/
    |-- README.md
    |-- coupled_state_sim.py
    `-- results/
        |-- cp_consistency_aggregate.csv
        |-- cp_consistency_independent_check.csv
        |-- cp_consistency_pairwise.csv
        |-- cp_consistency_per_turn.csv
        |-- cp_consistency_report.json
        `-- cp_consistency_summary.csv

Reproduction

cd <nrr-coupled-root>
bash scripts/run_repro_check.sh
bash scripts/build_current_manuscript.sh

Stable review-package entrypoints:

• bash scripts/run_repro_check.sh
• bash scripts/build_current_manuscript.sh
• bash scripts/verify_current_package.sh

License

CC BY 4.0. See LICENSE.

Contact

Kei Saito
Independent Researcher, Japan
ORCID: https://orcid.org/0009-0006-4715-9176
Email: kei.saito.research@gmail.com