A research corpus of publicly visible AI-generation artifact metadata — prompts, generation parameters (samplers, CFG, steps, ratios), model/LoRA stacks and engagement signals — aggregated across ~40 public content platforms for the study of generative-model usage patterns at scale. Archives are stored as sequential compressed blocks: - data/blk-.data — chunked, gzip-compressed, AES-256-GCM encrypted block files (segmented authenticated encryption, ~8 MiB segments). - index/manifest.data — encrypted manifest describing block grouping, checksums and provenance metadata. Archives are encrypted to keep the dataset licensing-neutral while we finalize a per-source redistribution review.…
Independent publisher
Elizabeth Adams
elizabethadams5935
Models in Library0
Datasets in Library2
Models on Hugging Face—
Followers—
Compare the published evidence
This profile connects the 2 model and dataset records shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.
Reported licenses: cc-by-4.0. Open a record to review its specific license and conditions.
Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.
Datasets
A durability-focused research corpus of AI-generated creative workflow media — node-graph session snapshots and their associated rendered assets (thumbnails, stills, and encoded previews) collected from public community galleries for generative content platforms. Media corpora for generative workflows are commonly stored on ephemeral CDNs with high link-rot. This corpus packages a snapshot of such assets in a sharded, integrity-checked, encrypted archival format suitable for durability/replication research and offline analysis of generative content characteristics (encoding formats, size distributions, thumbnail vs full-asset trade-offs). - shards/.data — encrypted (AES-256-CTR) tar…