Improve embedding methods, licenses, and descriptions

Hide internal QC/source-dataset fields, display linked upstream license/source terms, replace method placeholders from package metadata/READMEs, and remove redundant embedding-name prefixes.

app.js

@@ -33,6 +33,23 @@
     return div.innerHTML;
   };
 
+
+  const externalValue = (label, url) => {
+    const safeLabel = escapeHtml(label || "—");
+    const rawUrl = String(url || "").trim();
+
+    if (!/^https?:\/\//i.test(rawUrl)) {
+      return safeLabel;
+    }
+
+    return `
+      <a
+        href="${escapeHtml(rawUrl)}"
+        target="_blank"
+        rel="noopener noreferrer"
+      >${safeLabel}</a>`;
+  };
+
   const communityBadge = (item) => {
     if (item.community_contribution !== true) {
       return "";
@@ -215,21 +232,9 @@
         ${escapeHtml(item.algorithm || "—")}
       </p>
 
-      <p>
-        <b>QC:</b>
-        ${escapeHtml(item.qc_status || "legacy package")}
-      </p>
-
       <p>
         <b>License:</b>
-        ${escapeHtml(
-          item.license || "see package metadata"
-        )}
-      </p>
-
-      <p>
-        <b>Source Dataset:</b>
-        ${escapeHtml(item.source_dataset || "—")}
+        ${externalValue(item.license, item.license_url)}
       </p>
 
       ${contributorDetail(item)}

catalog.json

@@ -1,6 +1,6 @@
 [
   {
-    "algorithm": "signed log1p per summary feature, z-score across genes, PCA to 256 dimensions",
+    "algorithm": "Google DeepMind AlphaGenome API regulatory predictions. signed log1p per summary feature, z-score across genes, PCA to 256 dimensions.",
     "category": "Regulatory",
     "category_source": "manual",
     "community_contribution": false,
@@ -8,7 +8,7 @@
     "dataset_path": "data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256",
     "description": "Gene-level regulatory representation derived from AlphaGenome track-summary predictions around the transcription start site and reduced to 256 dimensions.",
     "id": "alphagenome_tss1mb_regulatory_tracksummary_pca256",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/alphagenome_tss1mb_regulatory_tracksummary_pca256/metadata.json",
     "modality": "regulatory genomics / AlphaGenome regulatory-track summary",
     "n_dims": 256,
@@ -18,18 +18,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/google-deepmind/alphagenome"
   },
   {
-    "algorithm": "Basenji2 PyTorch port",
+    "algorithm": "Basenji2 sequence-to-function model (checkpoint: Basenji2 PyTorch port). mean of central 3 output bins around TSS. log1p raw human central3 mean tracks, then dimension-wise z-score across genes.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 100.0,
     "dataset_path": "data/embeddings/basenji2_tss_131kb_human_central3",
-    "description": "Basenji2-TSS-131kb-human-central3-log1p-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "basenji2_tss_131kb_human_central3",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/basenji2_tss_131kb_human_central3/metadata.json",
     "modality": "regulatory DNA / Basenji2 TSS sequence model",
     "n_dims": 5313,
@@ -39,10 +40,11 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/14969593"
   },
   {
-    "algorithm": "dmis-lab/biobert-base-cased-v1.2",
+    "algorithm": "BioBERT-GeneProteinSummary (checkpoint: dmis-lab/biobert-base-cased-v1.2). Daniel locally. final hidden layer mean pooling over valid non-special, non-padding tokens.",
     "category": "Literature",
     "category_source": "manual",
     "community_contribution": false,
@@ -50,7 +52,7 @@
     "dataset_path": "data/embeddings/biobert_gene_protein_summary",
     "description": "NCBI gene card summaries plus UniProt protein summaries when available",
     "id": "biobert_gene_protein_summary",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/biobert_gene_protein_summary/metadata.json",
     "modality": "literature_text_gene_summary",
     "n_dims": 768,
@@ -60,18 +62,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/dmis-lab/biobert-base-cased-v1.2"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "BioConceptVec-CBOW.",
     "category": "Literature",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 84.79,
     "dataset_path": "data/embeddings/bioconceptvec_cbow",
-    "description": "BioConceptVec-CBOW-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "bioconceptvec_cbow",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/bioconceptvec_cbow/metadata.json",
     "modality": "biomedical_literature_text",
     "n_dims": 100,
@@ -81,18 +84,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/ncbi/BioConceptVec"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "BioConceptVec-fastText.",
     "category": "Literature",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 84.79,
     "dataset_path": "data/embeddings/bioconceptvec_fasttext",
-    "description": "BioConceptVec-fastText-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "bioconceptvec_fasttext",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/bioconceptvec_fasttext/metadata.json",
     "modality": "biomedical_literature_text",
     "n_dims": 100,
@@ -102,18 +106,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/ncbi/BioConceptVec"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "BioConceptVec-GloVe.",
     "category": "Literature",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 84.79,
     "dataset_path": "data/embeddings/bioconceptvec_glove",
-    "description": "BioConceptVec-GloVe-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "bioconceptvec_glove",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/bioconceptvec_glove/metadata.json",
     "modality": "biomedical_literature_text",
     "n_dims": 100,
@@ -123,18 +128,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/ncbi/BioConceptVec"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "BioConceptVec-SkipGram.",
     "category": "Literature",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 84.79,
     "dataset_path": "data/embeddings/bioconceptvec_skipgram",
-    "description": "BioConceptVec-SkipGram-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "bioconceptvec_skipgram",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/bioconceptvec_skipgram/metadata.json",
     "modality": "biomedical_literature_text",
     "n_dims": 100,
@@ -144,18 +150,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/ncbi/BioConceptVec"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "CMap/LINCS L1000 consensus gene signatures. mean 978-dimensional L1000 response vector across signatures and cell lines per perturbed gene. GCTX matrix interpreted as CGS signatures x 978 landmark response genes. Signatures were filtered to pert_type trt_sh.cgs, averaged by perturbed gene symbol across cell lines/signatures, then landmark response gene columns were z-scored across mapped perturbed genes.",
     "category": "Expression",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 20.25,
     "dataset_path": "data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore",
-    "description": "CMap-GSE106127-CGS-mean-landmark978-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "cmap_gse106127_cgs_mean_landmark978_zscore",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/cmap_gse106127_cgs_mean_landmark978_zscore/metadata.json",
     "modality": "perturbational transcriptomics / L1000 consensus genetic perturbation signatures",
     "n_dims": 978,
@@ -165,18 +172,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE106127"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "deepNF: deep network fusion for protein function prediction. Daniel locally from STRING v12 channel networks.",
     "category": "Network",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 94.89,
     "dataset_path": "data/embeddings/deepnf_string_v12_ppmi_svd_fusion",
-    "description": "deepNF-STRING-v12-PPMI-SVD-fusion-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "deepnf_string_v12_ppmi_svd_fusion",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/deepnf_string_v12_ppmi_svd_fusion/metadata.json",
     "modality": "STRING network multimodal sparse PPMI/SVD fusion",
     "n_dims": 256,
@@ -186,18 +194,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://string-db.org/cgi/download?species_text=Homo+sapiens"
   },
   {
-    "algorithm": "DeepSEA/predict from Kipoi",
+    "algorithm": "DeepSEA sequence-to-function model (checkpoint: DeepSEA/predict from Kipoi).",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 100.0,
     "dataset_path": "data/embeddings/deepsea_tss_1kb_epigenomic_probs",
-    "description": "DeepSEA-TSS-1kb-epigenomic-probabilities-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "deepsea_tss_1kb_epigenomic_probs",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/deepsea_tss_1kb_epigenomic_probs/metadata.json",
     "modality": "regulatory DNA / promoter-TSS sequence model",
     "n_dims": 919,
@@ -207,10 +216,11 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://kipoi.org/models/DeepSEA/predict/"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "DepMap CRISPR gene-effect matrix. Gene-effect matrix transposed to genes x models; missing values filled by source gene mean; model/cell-line features standardized across mapped genes; PCA with randomized SVD.",
     "category": "Functional Genomics",
     "category_source": "manual",
     "community_contribution": false,
@@ -218,7 +228,7 @@
     "dataset_path": "data/embeddings/depmap_crispr_gene_effect_pca256",
     "description": "Gene representation derived from DepMap CRISPR gene-effect profiles across cancer cell-line models, standardized and reduced to 256 dimensions using PCA.",
     "id": "depmap_crispr_gene_effect_pca256",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/depmap_crispr_gene_effect_pca256/metadata.json",
     "modality": "functional genomics / CRISPR dependency",
     "n_dims": 256,
@@ -228,18 +238,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://depmap.org/portal/download/"
   },
   {
-    "algorithm": "DNABERT-2 117M",
+    "algorithm": "DNABERT-2 genome language model (checkpoint: DNABERT-2 117M). Mean pooling over final-layer non-special token embeddings. Long CDS sequences split into nucleotide chunks, then length-weighted mean of chunk-level embeddings.",
     "category": "DNA",
     "category_source": "manual",
     "community_contribution": false,
     "coverage_pct": 99.94,
     "dataset_path": "data/embeddings/dnabert2_117m_cds",
-    "description": "dnabert2_117m_cds_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "dnabert2_117m_cds",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/dnabert2_117m_cds/metadata.json",
     "modality": "canonical protein-coding CDS DNA sequence",
     "n_dims": 768,
@@ -249,18 +260,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/zhihan1996/DNABERT-2-117M"
   },
   {
-    "algorithm": "Enformer",
+    "algorithm": "Enformer sequence-to-function model. mean of central 3 human output bins around TSS. log1p raw human central3 mean tracks, then dimension-wise z-score across genes.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 100.0,
     "dataset_path": "data/embeddings/enformer_tss_393kb_human_central3",
-    "description": "Enformer-TSS-393kb-human-central3-log1p-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "enformer_tss_393kb_human_central3",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/enformer_tss_393kb_human_central3/metadata.json",
     "modality": "regulatory DNA / long-range TSS sequence model",
     "n_dims": 5313,
@@ -270,18 +282,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://tfhub.dev/deepmind/enformer/1"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "ESM2 Transformer (checkpoint: esm2_t33_650M_UR50D, as reported by Zhong). Zhong et al. mean pooling over amino-acid token embeddings, as reported by Zhong.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.0,
     "dataset_path": "data/embeddings/esm2",
-    "description": "ESM2-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "esm2",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/esm2/metadata.json",
     "modality": "protein_sequence",
     "n_dims": 1280,
@@ -291,10 +304,11 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/16764517"
   },
   {
-    "algorithm": "FRoGS ARCHS4 gene embedding",
+    "algorithm": "FRoGS-ARCHS4-256 (checkpoint: FRoGS ARCHS4 gene embedding).",
     "category": "Functional Genomics",
     "category_source": "manual",
     "community_contribution": false,
@@ -302,7 +316,7 @@
     "dataset_path": "data/embeddings/frogs_archs4_256",
     "description": "FRoGS gene representation incorporating functional relationships learned using large-scale ARCHS4 gene-expression information.",
     "id": "frogs_archs4_256",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/frogs_archs4_256/metadata.json",
     "modality": "functional gene representation / ARCHS4 expression",
     "n_dims": 256,
@@ -312,18 +326,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/chenhcs/FRoGS"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "Gene2Vec.",
     "category": "Expression",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 89.15,
     "dataset_path": "data/embeddings/gene2vec",
-    "description": "Gene2Vec-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "gene2vec",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/gene2vec/metadata.json",
     "modality": "coexpression_expression",
     "n_dims": 200,
@@ -333,18 +348,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/jingcheng-du/Gene2vec"
   },
   {
-    "algorithm": "none; static learned token embedding weight extracted directly from checkpoint",
+    "algorithm": "GeneCompass Base, 12-layer knowledge-informed cross-species foundation model. Daniel locally by extracting checkpoint embedding weights; no training or inference. none; static learned token embedding weight extracted directly from checkpoint.",
     "category": "Single-cell",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 82.49,
     "dataset_path": "data/embeddings/genecompass_base_gene_token",
-    "description": "genecompass_base_gene_token_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "genecompass_base_gene_token",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/genecompass_base_gene_token/metadata.json",
     "modality": "gene expression single-cell foundation model / knowledge-informed gene-token embedding",
     "n_dims": 768,
@@ -354,18 +370,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/xCompass-AI/GeneCompass"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "Geneformer V2 316M input gene-token embedding layer. Extracted locally from official Geneformer model input gene-token embedding layer. none; static learned token embedding weight extracted directly from checkpoint.",
     "category": "Single-cell",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 99.55,
     "dataset_path": "data/embeddings/geneformer_v2_316m",
-    "description": "Geneformer-V2-316M-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "geneformer_v2_316m",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/geneformer_v2_316m/metadata.json",
     "modality": "single_cell_expression_foundation_model",
     "n_dims": 1152,
@@ -375,18 +392,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/ctheodoris/Geneformer"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "GenePT-ADA.",
     "category": "Literature",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 95.03,
     "dataset_path": "data/embeddings/genept_ada",
-    "description": "GenePT-ADA-v1.1-conservative. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "genept_ada_conservative",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/genept_ada/metadata.json",
     "modality": "biomedical_literature_text",
     "n_dims": 1536,
@@ -396,18 +414,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/10833191"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "GenePT-Model3-GeneProtein.",
     "category": "Literature",
     "category_source": "manual",
     "community_contribution": false,
     "coverage_pct": 95.05,
     "dataset_path": "data/embeddings/genept_model3_gene_protein",
-    "description": "GenePT-Model3-GeneProtein-v1.1-conservative. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "genept_model3_gene_protein_conservative",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/genept_model3_gene_protein/metadata.json",
     "modality": "biomedical_literature_text_plus_protein_sequence_text",
     "n_dims": 3072,
@@ -417,18 +436,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/10833191"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "GTEx RNA-seq median tissue expression matrix. TPM values were transformed with log1p, then each tissue column was z-scored across mapped genes.",
     "category": "Expression",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 96.3,
     "dataset_path": "data/embeddings/gtex_tissue_median_log1p_zscore",
-    "description": "GTEx-tissue-median-log1p-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "gtex_tissue_median_log1p_zscore",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/gtex_tissue_median_log1p_zscore/metadata.json",
     "modality": "bulk RNA expression / normal tissue expression",
     "n_dims": 68,
@@ -438,10 +458,11 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://gtexportal.org/home/datasets"
   },
   {
-    "algorithm": "helical-ai/helix-mRNA",
+    "algorithm": "Helix-mRNA (checkpoint: helical-ai/helix-mRNA). Daniel locally by model inference. final token embeddings from official Helix wrapper, mean pooled over non-special sequence tokens; long transcripts chunked and token-count weighted.",
     "category": "RNA",
     "category_source": "manual",
     "community_contribution": false,
@@ -449,7 +470,7 @@
     "dataset_path": "data/embeddings/helix_mrna_full_transcript",
     "description": "Gene-level mature-transcript representation generated with Helix-mRNA and mean-pooled across non-special sequence tokens.",
     "id": "helix_mrna_full_transcript",
-    "license": "not recorded in package metadata",
+    "license": "model_license_on_model_card: cc-by-nc-sa-4.0",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/helix_mrna_full_transcript/metadata.json",
     "modality": "RNA language model / mature mRNA transcript",
     "n_dims": 256,
@@ -459,18 +480,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/helical-ai/helix-mRNA"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "Hetionet v1.0 heterogeneous biomedical knowledge graph. Built sparse gene x typed-neighbor-feature count matrix; collapsed duplicate gene-feature edges by summation; applied sublinear TF-IDF weighting; reduced with randomized TruncatedSVD; z-scored SVD dimensions across source genes; mapped to master by Entrez ID.",
     "category": "Knowledge",
     "category_source": "manual",
     "community_contribution": false,
     "coverage_pct": 93.48,
     "dataset_path": "data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256",
-    "description": "Hetionet-nonPPI-biomedical-context-TFIDF-SVD256-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "hetionet_nonppi_biomedical_context_tfidf_svd256",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/hetionet_nonppi_biomedical_context_tfidf_svd256/metadata.json",
     "modality": "heterogeneous biomedical knowledge graph / non-PPI gene context",
     "n_dims": 256,
@@ -480,18 +502,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/268568"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "HPA normal tissue immunohistochemistry annotations. maximum ordinal protein-expression level across annotated cell types within each tissue. HPA Level encoded as Not detected=0, Low=1, Medium=2, High=3. For each gene and tissue, maximum level across annotated cell types was used. Tissue columns were z-scored across mapped genes.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 66.78,
     "dataset_path": "data/embeddings/hpa_normal_ihc_tissue_level_zscore",
-    "description": "HPA-normal-IHC-tissue-level-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "hpa_normal_ihc_tissue_level_zscore",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/hpa_normal_ihc_tissue_level_zscore/metadata.json",
     "modality": "protein expression / normal tissue immunohistochemistry",
     "n_dims": 59,
@@ -501,18 +524,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://www.proteinatlas.org/humanproteome/tissue/data"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "HPA subcellular immunofluorescence annotations. Main location, Additional location, and Extracellular location were parsed as semicolon-separated labels. A binary multi-hot location vector was created per gene, duplicate genes were collapsed by max, then location columns were z-scored across mapped genes.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 67.58,
     "dataset_path": "data/embeddings/hpa_subcellular_location_multihot_zscore",
-    "description": "HPA-subcellular-location-multihot-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "hpa_subcellular_location_multihot_zscore",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/hpa_subcellular_location_multihot_zscore/metadata.json",
     "modality": "protein subcellular localization",
     "n_dims": 50,
@@ -522,7 +546,8 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://www.proteinatlas.org/humanproteome/subcellular/data"
   },
   {
     "algorithm": "Deterministic Gaussian synthetic fixture",
@@ -530,7 +555,7 @@
     "category": "Protein",
     "category_source": "community_submission",
     "community_contribution": true,
-    "contributor_affiliation": "ETH Z\u00fcrich",
+    "contributor_affiliation": "ETH Zürich",
     "contributor_hf_username": "thecosmicboy",
     "contributor_name": "Daniel Tabares",
     "coverage_pct": 7.5,
@@ -588,7 +613,7 @@
     "validation_path": "data/community_contributions/hf_portal_root_upload_smoke_20260729_1328_v1"
   },
   {
-    "algorithm": "final hidden-state mean pooling over non-special nucleotide tokens; long CDS split into chunks and token-count weighted",
+    "algorithm": "HyenaDNA medium 160k sequence-length checkpoint (checkpoint: LongSafari/hyenadna-medium-160k-seqlen-hf). Daniel locally by model inference. final hidden-state mean pooling over non-special nucleotide tokens; long CDS split into chunks and token-count weighted.",
     "category": "DNA",
     "category_source": "manual",
     "community_contribution": false,
@@ -596,7 +621,7 @@
     "dataset_path": "data/embeddings/hyenadna_medium_160k_cds",
     "description": "Gene-level canonical-CDS representation generated with the HyenaDNA medium 160k sequence-length model.",
     "id": "hyenadna_medium_160k_cds",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/hyenadna_medium_160k_cds/metadata.json",
     "modality": "DNA language model / canonical coding sequence",
     "n_dims": 256,
@@ -606,10 +631,11 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/LongSafari/hyenadna-medium-160k-seqlen-hf"
   },
   {
-    "algorithm": "mean across all non-global tissue/cell contexts per Entrez gene",
+    "algorithm": "Mahi context-aware gene embedding model. mean across all non-global tissue/cell contexts per Entrez gene.",
     "category": "Multimodal",
     "category_source": "manual",
     "community_contribution": false,
@@ -617,7 +643,7 @@
     "dataset_path": "data/embeddings/mahi_all_contexts_mean",
     "description": "Mahi gene representation formed by averaging embeddings across all available non-global tissue and cell contexts.",
     "id": "mahi_all_contexts_mean",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/mahi_all_contexts_mean/metadata.json",
     "modality": "multimodal tissue-aware graph representation / mean across non-global contexts",
     "n_dims": 3666,
@@ -627,10 +653,11 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/datasets/anushaggs/Mahi_gene_embeddings"
   },
   {
-    "algorithm": "global context only",
+    "algorithm": "Mahi context-aware gene embedding model. global context only.",
     "category": "Multimodal",
     "category_source": "manual",
     "community_contribution": false,
@@ -638,7 +665,7 @@
     "dataset_path": "data/embeddings/mahi_global",
     "description": "Global-context Mahi gene representation integrating regulatory, protein and network-derived molecular information.",
     "id": "mahi_global",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/mahi_global/metadata.json",
     "modality": "multimodal tissue-aware graph representation / global context",
     "n_dims": 3666,
@@ -648,18 +675,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/datasets/anushaggs/Mahi_gene_embeddings"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "Mashup-STRING.",
     "category": "Network",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 81.26,
     "dataset_path": "data/embeddings/mashup_string",
-    "description": "Mashup-STRING-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "mashup_string",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/mashup_string/metadata.json",
     "modality": "network_ppi",
     "n_dims": 800,
@@ -669,18 +697,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://mashup.csail.mit.edu/"
   },
   {
-    "algorithm": "esm_msa1b_t12_100M_UR50S",
+    "algorithm": "MSA Transformer (checkpoint: esm_msa1b_t12_100M_UR50S). Daniel locally. mean residue representation from layer 12; long proteins split into non-overlapping windows and length-weighted mean pooled.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 96.89,
     "dataset_path": "data/embeddings/msa_transformer_depth1_uniprot_human",
-    "description": "MSATransformer-depth1-UniProt-human-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "msa_transformer_depth1_uniprot_human",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/msa_transformer_depth1_uniprot_human/metadata.json",
     "modality": "protein_sequence_msa_transformer_depth1",
     "n_dims": 768,
@@ -690,18 +719,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/facebookresearch/esm"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "NEWT component embedding (checkpoint: not applicable). NEWT authors; processed locally into the project standard format. precomputed component embedding table.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 80.93,
     "dataset_path": "data/embeddings/newt_archs4_256",
-    "description": "NEWT-ARCHS4-256-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "newt_archs4_256",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/newt_archs4_256/metadata.json",
     "modality": "multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT",
     "n_dims": 256,
@@ -711,18 +741,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/16764517"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "NEWT component embedding (checkpoint: not applicable). NEWT authors; processed locally into the project standard format. precomputed component embedding table.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 87.07,
     "dataset_path": "data/embeddings/newt_cellnet",
-    "description": "NEWT-CellNet-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "newt_cellnet",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/newt_cellnet/metadata.json",
     "modality": "multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT",
     "n_dims": 128,
@@ -732,18 +763,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/16764517"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "NEWT component embedding (checkpoint: not applicable). NEWT authors; processed locally into the project standard format. precomputed component embedding table.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 68.61,
     "dataset_path": "data/embeddings/newt_go_256",
-    "description": "NEWT-GO-256-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "newt_go_256",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/newt_go_256/metadata.json",
     "modality": "multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT",
     "n_dims": 256,
@@ -753,18 +785,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/16764517"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "NEWT component embedding (checkpoint: not applicable). NEWT authors; processed locally into the project standard format. precomputed component embedding table.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 68.61,
     "dataset_path": "data/embeddings/newt_go_graph",
-    "description": "NEWT-GO-graph-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "newt_go_graph",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/newt_go_graph/metadata.json",
     "modality": "multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT",
     "n_dims": 256,
@@ -774,18 +807,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/16764517"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "NEWT component embedding (checkpoint: not applicable). NEWT authors; processed locally into the project standard format. precomputed component embedding table.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 76.74,
     "dataset_path": "data/embeddings/newt_msigdb_bundle",
-    "description": "NEWT-MSigDB-bundle-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "newt_msigdb_bundle",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/newt_msigdb_bundle/metadata.json",
     "modality": "multimodal functional/regulatory/co-expression/pathway gene embedding component from NEWT",
     "n_dims": 256,
@@ -795,18 +829,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://zenodo.org/records/16764517"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "node2vec: scalable feature learning for networks. Daniel locally from consensus.dat using reproducible random walks and Word2Vec skip-gram.",
     "category": "Network",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.34,
     "dataset_path": "data/embeddings/node2vec_consensus_ppi",
-    "description": "Node2Vec-consensus-PPI-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "node2vec_consensus_ppi",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/node2vec_consensus_ppi/metadata.json",
     "modality": "PPI network embedding",
     "n_dims": 128,
@@ -816,10 +851,11 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/ylaboratory/gene-embedding-benchmarks"
   },
   {
-    "algorithm": "InstaDeepAI/nucleotide-transformer-v2-500m-multi-species",
+    "algorithm": "Nucleotide Transformer v2 (checkpoint: InstaDeepAI/nucleotide-transformer-v2-500m-multi-species). Daniel locally by model inference. final hidden layer mean pooling over non-special, non-padding tokens; chunk-weighted by token count for long CDS.",
     "category": "DNA",
     "category_source": "manual",
     "community_contribution": false,
@@ -827,7 +863,7 @@
     "dataset_path": "data/embeddings/nt_v2_500m_multispecies_cds",
     "description": "Gene-level canonical-CDS representation generated with the 500M-parameter multispecies Nucleotide Transformer v2 checkpoint.",
     "id": "nt_v2_500m_multispecies_cds",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/nt_v2_500m_multispecies_cds/metadata.json",
     "modality": "DNA language model / canonical coding sequence",
     "n_dims": 1024,
@@ -837,18 +873,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/InstaDeepAI/nucleotide-transformer-v2-500m-multi-species"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "OPA2Vec. Daniel locally using an OPA2Vec-derived Python 3 workflow.",
     "category": "Knowledge",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 92.85,
     "dataset_path": "data/embeddings/opa2vec_goa_human",
-    "description": "OPA2Vec-GOA-human-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "opa2vec_goa_human",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/opa2vec_goa_human/metadata.json",
     "modality": "Gene Ontology / ontology-annotation semantic embedding",
     "n_dims": 200,
@@ -858,10 +895,11 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/bio-ontology-research-group/opa2vec"
   },
   {
-    "algorithm": "Orthrus base 4-track",
+    "algorithm": "Orthrus mature RNA foundation model (checkpoint: Orthrus base 4-track). Orthrus representation() sequence-level embedding. For transcripts longer than max_full_len, length-weighted mean of chunk-level Orthrus embeddings.",
     "category": "RNA",
     "category_source": "manual",
     "community_contribution": false,
@@ -869,7 +907,7 @@
     "dataset_path": "data/embeddings/orthrus_base_4track_full_transcript",
     "description": "Mature-transcript representation generated with the Orthrus base four-track RNA foundation model.",
     "id": "orthrus_base_4track_full_transcript",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/orthrus_base_4track_full_transcript/metadata.json",
     "modality": "RNA foundation model / mature mRNA transcript",
     "n_dims": 256,
@@ -879,18 +917,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/quietflamingo/orthrus-base-4-track"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "PrimeKG heterogeneous biomedical knowledge graph. Built sparse gene x typed-neighbor-feature count matrix; collapsed duplicate gene-feature edges by summation; applied sublinear TF-IDF weighting; reduced with randomized TruncatedSVD; z-scored SVD dimensions across source genes; mapped to master by Entrez ID.",
     "category": "Knowledge",
     "category_source": "manual",
     "community_contribution": false,
     "coverage_pct": 95.62,
     "dataset_path": "data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256",
-    "description": "PrimeKG-nonPPI-biomedical-context-TFIDF-SVD256-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "primekg_nonppi_biomedical_context_tfidf_svd256",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/primekg_nonppi_biomedical_context_tfidf_svd256/metadata.json",
     "modality": "heterogeneous biomedical knowledge graph / non-PPI gene context",
     "n_dims": 256,
@@ -900,18 +939,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/IXA7BM"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "AAC.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_aac",
-    "description": "probe_aac_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_aac",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_aac/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 20,
@@ -921,18 +961,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "ALBERT.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.82,
     "dataset_path": "data/embeddings/probe_albert",
-    "description": "probe_albert_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_albert",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_albert/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 4096,
@@ -942,18 +983,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "APAAC.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.9,
     "dataset_path": "data/embeddings/probe_apaac",
-    "description": "probe_apaac_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_apaac",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_apaac/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 80,
@@ -963,18 +1005,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "BERT-BFD.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.82,
     "dataset_path": "data/embeddings/probe_bert_bfd",
-    "description": "probe_bert_bfd_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_bert_bfd",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_bert_bfd/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 1024,
@@ -984,18 +1027,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "BERT-PFAM.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.86,
     "dataset_path": "data/embeddings/probe_bert_pfam",
-    "description": "probe_bert_pfam_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_bert_pfam",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_bert_pfam/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 768,
@@ -1005,18 +1049,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "BLAST.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_blast",
-    "description": "probe_blast_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_blast",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_blast/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 20421,
@@ -1026,18 +1071,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "CPC-PROT.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.82,
     "dataset_path": "data/embeddings/probe_cpc_prot",
-    "description": "probe_cpc_prot_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_cpc_prot",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_cpc_prot/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 512,
@@ -1047,18 +1093,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "ESMB1.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.82,
     "dataset_path": "data/embeddings/probe_esmb1",
-    "description": "probe_esmb1_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_esmb1",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_esmb1/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 1280,
@@ -1068,18 +1115,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "GENE2VEC.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 87.4,
     "dataset_path": "data/embeddings/probe_gene2vec_uniprot",
-    "description": "probe_gene2vec_uniprot_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_gene2vec_uniprot",
-    "license": "not recorded in package metadata",
+    "license": "PROBE source / license information",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_gene2vec_uniprot/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 200,
@@ -1089,18 +1137,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/tuncadogan/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "HMMER.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_hmmer",
-    "description": "probe_hmmer_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_hmmer",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_hmmer/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 20421,
@@ -1110,18 +1159,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "KSEP.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.55,
     "dataset_path": "data/embeddings/probe_ksep",
-    "description": "probe_ksep_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_ksep",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_ksep/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 400,
@@ -1131,18 +1181,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 23
+    "zero_vectors": 23,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "LEARNED-VEC.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_learned_vec",
-    "description": "probe_learned_vec_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_learned_vec",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_learned_vec/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 64,
@@ -1152,18 +1203,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "MUT2VEC.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 86.8,
     "dataset_path": "data/embeddings/probe_mut2vec",
-    "description": "probe_mut2vec_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_mut2vec",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_mut2vec/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 300,
@@ -1173,18 +1225,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "PFAM.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_pfam",
-    "description": "probe_pfam_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_pfam",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_pfam/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 6227,
@@ -1194,18 +1247,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 1532
+    "zero_vectors": 1532,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "PROTVEC.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_protvec",
-    "description": "probe_protvec_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_protvec",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_protvec/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 100,
@@ -1215,18 +1269,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "SEQVEC.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_seqvec",
-    "description": "probe_seqvec_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_seqvec",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_seqvec/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 1024,
@@ -1236,18 +1291,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "T5.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.82,
     "dataset_path": "data/embeddings/probe_t5",
-    "description": "probe_t5_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_t5",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_t5/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 1024,
@@ -1257,18 +1313,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "TCGA-EMBEDDING.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 90.67,
     "dataset_path": "data/embeddings/probe_tcga_embedding",
-    "description": "probe_tcga_embedding_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_tcga_embedding",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_tcga_embedding/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 50,
@@ -1278,18 +1335,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "UNIREP.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 93.91,
     "dataset_path": "data/embeddings/probe_unirep",
-    "description": "probe_unirep_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_unirep",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_unirep/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 5700,
@@ -1299,18 +1357,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "XLNET.",
     "category": "Protein",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 91.82,
     "dataset_path": "data/embeddings/probe_xlnet",
-    "description": "probe_xlnet_v1_1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "probe_xlnet",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/probe_xlnet/metadata.json",
     "modality": "protein sequence / protein representation from PROBE human UniProt archive",
     "n_dims": 1024,
@@ -1320,10 +1379,11 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/spaces/HUBioDataLab/PROBE"
   },
   {
-    "algorithm": "microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext",
+    "algorithm": "PubMedBERT-GeneProteinSummary (checkpoint: microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext). Daniel locally. final hidden layer mean pooling over valid non-special, non-padding tokens.",
     "category": "Literature",
     "category_source": "manual",
     "community_contribution": false,
@@ -1331,7 +1391,7 @@
     "dataset_path": "data/embeddings/pubmedbert_gene_protein_summary",
     "description": "NCBI gene card summaries plus UniProt protein summaries when available",
     "id": "pubmedbert_gene_protein_summary",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/pubmedbert_gene_protein_summary/metadata.json",
     "modality": "literature_text_gene_summary",
     "n_dims": 768,
@@ -1341,10 +1401,11 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext"
   },
   {
-    "algorithm": "RNA-FM / rna_fm_t12",
+    "algorithm": "RNA-FM / rna_fm_t12 RNA language model. Final layer 12 mean token pooling over real RNA tokens only; BOS/EOS excluded; long transcripts split into chunks and chunk means length-weighted.",
     "category": "RNA",
     "category_source": "manual",
     "community_contribution": false,
@@ -1352,7 +1413,7 @@
     "dataset_path": "data/embeddings/rnafm_full_transcript",
     "description": "Mature-transcript representation generated from the final layer of RNA-FM using length-weighted token pooling.",
     "id": "rnafm_full_transcript",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/rnafm_full_transcript/metadata.json",
     "modality": "RNA language model / mature mRNA transcript",
     "n_dims": 640,
@@ -1362,18 +1423,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/ml4bio/RNA-FM"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "scFoundation / xTrimoGene (checkpoint: models.ckpt). Extracted locally from official scFoundation gene sub-checkpoint model.pos_emb.weight. fixed checkpoint embedding table; no cell/context pooling.",
     "category": "Single-cell",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 94.85,
     "dataset_path": "data/embeddings/scfoundation_gene_pos",
-    "description": "scFoundation-gene-pos-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "scfoundation_gene_pos",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/scfoundation_gene_pos/metadata.json",
     "modality": "single_cell_expression_foundation_model / learned gene identity-position embedding",
     "n_dims": 768,
@@ -1383,18 +1445,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/biomap-research/scFoundation"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "scGPT whole-human input gene-token embedding layer. Extracted locally from official scGPT whole-human model input gene-token embedding layer. none; static learned token embedding weight extracted directly from checkpoint.",
     "category": "Single-cell",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 95.93,
     "dataset_path": "data/embeddings/scgpt_whole_human",
-    "description": "scGPT-whole-human-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "scgpt_whole_human",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/scgpt_whole_human/metadata.json",
     "modality": "single_cell_expression_foundation_model",
     "n_dims": 512,
@@ -1404,10 +1467,11 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/bowang-lab/scGPT"
   },
   {
-    "algorithm": "scGPT pan-cancer gene-token embedding",
+    "algorithm": "scGPT pan-cancer checkpoint (checkpoint: scGPT pan-cancer gene-token embedding). local extraction from official checkpoint input gene-token embedding table. none; static learned token embedding weight extracted directly from checkpoint.",
     "category": "Single-cell",
     "category_source": "manual",
     "community_contribution": false,
@@ -1415,7 +1479,7 @@
     "dataset_path": "data/embeddings/scgpt_pancancer",
     "description": "Gene-token representation derived from a pan-cancer scGPT single-cell transcriptomic model.",
     "id": "scgpt_pancancer",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/scgpt_pancancer/metadata.json",
     "modality": "single-cell transcriptomic foundation model / pan-cancer gene-token embedding",
     "n_dims": 512,
@@ -1425,18 +1489,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/bowang-lab/scGPT"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "scPRINT (checkpoint: medium-v1.5.ckpt). Extracted locally from official scPRINT medium-v1.5 checkpoint. fixed checkpoint embedding table; no cell/context pooling.",
     "category": "Single-cell",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 99.49,
     "dataset_path": "data/embeddings/scprint_medium_v1_5_gene",
-    "description": "scPRINT-medium-v1.5-gene-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "scprint_medium_v1_5_gene",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/scprint_medium_v1_5_gene/metadata.json",
     "modality": "single_cell_expression_foundation_model / learned gene-token embedding",
     "n_dims": 256,
@@ -1446,18 +1511,19 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://huggingface.co/jkobject/scPRINT"
   },
   {
-    "algorithm": "SpliceAI",
+    "algorithm": "SpliceAI splice-site prediction model. aggregate canonical splice-site profiles to one fixed-size vector per gene.",
     "category": "Regulatory",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 100.0,
     "dataset_path": "data/embeddings/spliceai_canonical_splice_site_profile",
-    "description": "SpliceAI-canonical-splice-site-profile-zscore-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "spliceai_canonical_splice_site_profile",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/spliceai_canonical_splice_site_profile/metadata.json",
     "modality": "splicing sequence model / canonical transcript splice-site profiles",
     "n_dims": 606,
@@ -1467,18 +1533,19 @@
     "qc_status": "processed_qc_passed",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/Illumina/SpliceAI"
   },
   {
-    "algorithm": "see package metadata",
+    "algorithm": "Universal Cell Embeddings (UCE) (checkpoint: 33l_8ep_1024t_1280.torch). Extracted locally from official UCE all_tokens and 33-layer checkpoint gene encoder. UCE gene encoder projection: Linear(5120->1280), ReLU, LayerNorm.",
     "category": "Single-cell",
     "category_source": "heuristic",
     "community_contribution": false,
     "coverage_pct": 94.95,
     "dataset_path": "data/embeddings/uce_33l",
-    "description": "UCE-33L-v1.1. Core embedding package from embeddingsSSRF/gene_embeddings.",
+    "description": "Core embedding package from embeddingsSSRF/gene_embeddings.",
     "id": "uce_33l",
-    "license": "not recorded in package metadata",
+    "license": "Original source / license terms",
     "metadata_url": "https://huggingface.co/datasets/embeddingsSSRF/gene_embeddings/blob/326b257a20e73df337ece40bb82401e277851f5b/data/embeddings/uce_33l/metadata.json",
     "modality": "single_cell_expression_foundation_model / protein-informed gene-token embedding",
     "n_dims": 1280,
@@ -1488,6 +1555,7 @@
     "qc_status": "legacy_core_package",
     "source_dataset": "embeddingsSSRF/gene_embeddings",
     "source_revision": "326b257a20e73df337ece40bb82401e277851f5b",
-    "zero_vectors": 0
+    "zero_vectors": 0,
+    "license_url": "https://github.com/snap-stanford/UCE"
   }
 ]