Armiger Plugin #2 — Swarm & Release-Group Empirical Bayes Priors (armiger-swarm-priors)

Zero-dependency Empirical Bayes prior engine implementing Armiger's 4 strictly separated swarm signal pathways, mandatory modality dropout, and public-tracker passkey sanitization.
Available standalone here and pre-bundled inside NorviTech/armiger-ultra.

The 4 Signal Pathways (swarm_priors.py)

1. Way 1 — Active Teacher Sampling & Queue Priority Weighting (compute_sampling_priority_weight): Prioritizes high-uncertainty and high-demand swarms in Bitmagnet and self-hosted indexer ingestion queues.
2. Way 2 — Public Tracker Domain & Release-Group Bayesian Priors (extract_way2_source_features): Extracts shrinkage-smoothed release-group, codec, resolution, and public-tracker domain affinities with Mandatory Modality Dropout ($p \in [0.50, 0.70]$) and strict invariance to seeders == 0.
3. Way 3 — Swarm-vs-Content Anomaly Features (extract_way3_anomaly_features): Computes seeder inflation $z$-scores, size-vs-resolution mismatches, and executable/shortcut trap indicators for junk / malware detection.
4. Way 4 — Orthogonal Swarm Viability & Stage-2 Grab Ranking (compute_swarm_viability): Produces a calibrated [0.0, 1.0] download viability multiplier (86.64% standalone formula AUROC; 96.83% multi-source corpus / 92.42% public-tracker AUROC when paired with armiger-swarm-survival) without ever altering the predicted content category.

Quickstart

from swarm_priors import (
    extract_way2_source_features,
    extract_way3_anomaly_features,
    compute_swarm_viability,
)

w2 = extract_way2_source_features(
    title="[OpenAnime] Spring - 04 (1080p) [E4A1B9C2].mkv",
    sources=["torrent.ubuntu.com", "bt1.archive.org"],
    total_size_bytes=1_420_000_000,
    files=[{"path": "[OpenAnime] Spring - 04 (1080p).mkv", "size": 1_420_000_000}],
)
w3 = extract_way3_anomaly_features(
    title="[OpenAnime] Spring - 04 (1080p) [E4A1B9C2].mkv",
    seeders=185,
    leechers=12,
    completed=940,
    total_size_bytes=1_420_000_000,
    files=[{"path": "[OpenAnime] Spring - 04 (1080p).mkv", "size": 1_420_000_000}],
)
print(compute_swarm_viability(seeders=185, leechers=12, completed=940, category="tv_show", way2_features=w2, way3_features=w3))

Part of the Armiger Modular Suite

This plugin is bundled inside NorviTech/armiger-ultra alongside all 6 official plugins (armiger-swarm-survival, armiger-swarm-priors, armiger-zero-cue-catalog, armiger-audio-guard, armiger-calibration-dht, and armiger-dht-db-sync) and is compatible with NorviTech/armiger (8.3 MB CPU INT8).

Intended Use & Legal Compliance

Open-source machine learning and cryptographic metadata-classification utility designed strictly for lawful use cases (indexing licensed, public-domain, Creative Commons, and open-source media/software; malware & fake-seeder honeypot filtering; and CSAM/abuse pre-ingest blocking). NorviTech does not condone, encourage, or support copyright infringement. See LEGAL.md and NOTICE.

---

Armiger Ultra · Armiger · Swarm Survival · Swarm Priors · Zero-Cue Catalog · Audio Guard · Calibration & DHT RSI · DHT DB Sync · norvitech.com