Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2025
|
| Online Access: | https://doi.org/10.5281/zenodo.17078519 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866902169513361408 |
|---|---|
| author | RS0136 |
| author_facet | RS0136 |
| contents | <h2>Summary</h2> <p>First stable release of the AIP Simulation Suite. This version generates all 14 paper figures (PNG) with matching LaTeX tables and CSV data, demonstrating the affine equivalence between Expected Free Energy (AIF/EFE) and pragmatic utility in RSA under assumptions A1–A6.</p> <h2>Highlights</h2> <ul> <li>Deterministic parallel execution (per-seed RNG + seed-sorted collation)</li> <li>14 figures + 14 paired LaTeX tables, ready to \input into your preprint</li> <li>Calibration, ablation (risk/ambiguity/IG), noise robustness, parameter recovery</li> <li>Model comparison (LL, BIC), ΔBIC-ready tables</li> <li>No seaborn, no custom colors; single-plot, publication-friendly figures</li> <li>Corpus used only to shape distributions; synthetic fallback if missing</li> </ul> <h2>Requirements</h2> <ul> <li>Python ≥ 3.9</li> <li>numpy, pandas, matplotlib, joblib</li> </ul> <h2>Install</h2> <p>pip install numpy pandas matplotlib joblib</p> <h2>Usage</h2> <p>python aip.py --input <folder-with-filteredCorpus.csv> --output <outdir> --cores <n> <br> [--trials-per-seed 1000] [--seeds 0-29]</p> <h2>Reproducibility</h2> <ul> <li>Default seeds: 0–29</li> <li>Results collated in deterministic seed order</li> <li>Verify stability by checksumming <code>output/data/per_seed_summary.csv</code> across repeated runs</li> </ul> <h2>Artifacts</h2> <ul> <li>output/figures/fig01.png ... fig14.png</li> <li>output/tables/fig01_table.tex ... fig14_table.tex</li> <li>output/data/*.csv, output/run_config.json</li> </ul> <h2>Notes</h2> <ul> <li>L0 uses probabilistic HSL semantics; bridge makes β ↔ 1/α explicit.</li> <li>If <code>filteredCorpus.csv</code> is absent, a small synthetic corpus is generated for smoke tests.</li> </ul> <h2>License</h2> <p>MIT</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17078519 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | RS0136/AIP: AIP Simulation Suite v1.0.0 — Reproducible Figures & LaTeX Tables for AIF ⇔ RSA RS0136 <h2>Summary</h2> <p>First stable release of the AIP Simulation Suite. This version generates all 14 paper figures (PNG) with matching LaTeX tables and CSV data, demonstrating the affine equivalence between Expected Free Energy (AIF/EFE) and pragmatic utility in RSA under assumptions A1–A6.</p> <h2>Highlights</h2> <ul> <li>Deterministic parallel execution (per-seed RNG + seed-sorted collation)</li> <li>14 figures + 14 paired LaTeX tables, ready to \input into your preprint</li> <li>Calibration, ablation (risk/ambiguity/IG), noise robustness, parameter recovery</li> <li>Model comparison (LL, BIC), ΔBIC-ready tables</li> <li>No seaborn, no custom colors; single-plot, publication-friendly figures</li> <li>Corpus used only to shape distributions; synthetic fallback if missing</li> </ul> <h2>Requirements</h2> <ul> <li>Python ≥ 3.9</li> <li>numpy, pandas, matplotlib, joblib</li> </ul> <h2>Install</h2> <p>pip install numpy pandas matplotlib joblib</p> <h2>Usage</h2> <p>python aip.py --input <folder-with-filteredCorpus.csv> --output <outdir> --cores <n> <br> [--trials-per-seed 1000] [--seeds 0-29]</p> <h2>Reproducibility</h2> <ul> <li>Default seeds: 0–29</li> <li>Results collated in deterministic seed order</li> <li>Verify stability by checksumming <code>output/data/per_seed_summary.csv</code> across repeated runs</li> </ul> <h2>Artifacts</h2> <ul> <li>output/figures/fig01.png ... fig14.png</li> <li>output/tables/fig01_table.tex ... fig14_table.tex</li> <li>output/data/*.csv, output/run_config.json</li> </ul> <h2>Notes</h2> <ul> <li>L0 uses probabilistic HSL semantics; bridge makes β ↔ 1/α explicit.</li> <li>If <code>filteredCorpus.csv</code> is absent, a small synthetic corpus is generated for smoke tests.</li> </ul> <h2>License</h2> <p>MIT</p> |
| title | RS0136/AIP: AIP Simulation Suite v1.0.0 — Reproducible Figures & LaTeX Tables for AIF ⇔ RSA |
| url | https://doi.org/10.5281/zenodo.17078519 |