Model type
Study-specific predictive method; this record is the paper-specific evaluated configuration.
Configuration
Best frozen single-cell foundation model
This study compares frozen single-cell foundation-model probes with a gene-expression PCA baseline for donor age.
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 2); Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1)How it worksEvaluated procedure (conceptual)
Conceptual input–method–output guide. Check the procedure text and linked evaluation for fitted components, additional inputs and exact settings.
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1)At a glance
Biological inputs
PBMC single-cell expression data with donor identities and age labels
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Methods/Statistical safeguards (paragraph 1); Methods/Datasets (paragraph 1)Outputs
Donor-age predictions and representation analyses
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Discussion/Limitations (paragraph 1); Methods/Statistical safeguards (paragraph 1)limited source coverage · Automated source review, 2026-09-16. All specifications and missing details
Evaluations and results
Release 2026-09-17-d277315f7d76 · 1 evaluation · 1 metric row. Different protocols are not a single leaderboard.
| Metric and finding | Coverage and uncertainty | Evidence |
|---|---|---|
| Best frozen single-cell foundation model: Donor-aware age-class prediction Configuration: Best frozen single-cell foundation modelTask: Donor-aware age-class predictionDataset: AIDA v2 PBMC cohort Same donor-aware splits and logistic-regression probe as expression PCA; text names Geneformer as best model on AIDA v2. Independent external evaluation · Evaluation metadata: needs review | ||
| 0.322 Balanced accuracy Unit: unitless · Direction: unknown | Uncertainty: ± 0.008 standard deviation Scored: Not reported · Eligible: Not reported | source checkedInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Table 2, AIDA v2 row, scFM BA ± SD column Source checking is not independent reproduction. |
How it works
How the evaluated method works
Frozen scGPT and Geneformer representations feed age probes. The conventional baseline uses 50 principal components of gene expression with regression under the same donor-aware splits.
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1)What was evaluated
The linked evaluation record identifies Best frozen single-cell foundation model: Donor-aware age-class prediction. Its dataset, split, adaptation and evidence origin remain attached to the reported results.
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · The named evaluation’s methods and comparison table; exact preserved evaluation IDs: evaluation-lit-b3-013Strengths and limitations
Strengths and considerations
- Donor-aware splitting and cell-composition matching probe important confounding mechanisms.
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Methods/Statistical safeguards (paragraph 1); Methods/Interpretability blocks (paragraph 1)
Limitations and conditions
- The “best frozen model” label is a selection across candidate probes, not a uniquely specified model checkpoint; PCA matches or exceeds its age-prediction performance in the reported cohorts.
Sources
Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1)
Profile review details
Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction.
Stable record: reported-model-148b613975b6ebSpecifications
Inputs, training, access and other details
Explanatory profile: limited source coverage · Automated source review, 2026-09-16. Review applies to the cited claims; unresolved fields are listed below. Numerical results retain their own review status.
| Property | Description and evidence |
|---|---|
| Model type | Study-specific predictive method; this record is the paper-specific evaluated configuration.SourcesInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) |
| Architecture / procedure | Frozen scGPT and Geneformer representations feed age probes. The conventional baseline uses 50 principal components of gene expression with regression under the same donor-aware splits.SourcesInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) |
| Biological inputs | PBMC single-cell expression data with donor identities and age labelsSourcesInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Methods/Statistical safeguards (paragraph 1); Methods/Datasets (paragraph 1) |
| Outputs | Donor-age predictions and representation analysesSourcesInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Discussion/Limitations (paragraph 1); Methods/Statistical safeguards (paragraph 1) |
| Parameters | This aggregate label does not identify a unique parameter count. The paper separately describes Geneformer V1-10M, V2-104M and V2-316M in its scaling analysis; those variants are not collapsed into one model identity. · Not reported in inspected sourcesSourcesInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Discussion / Robustness and control analyses; cross-model summary label versus Cross-Geneformer-size benchmark |
| Known versions / configuration | Best frozen single-cell foundation model is the comparison-table label; that label does not specify an immutable weight revision. · Not reported in inspected sourcesSourcesInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label. |
| Training data / fitting | Five PBMC cohorts containing approximately 4–5 million cells from about 2,000 donors; foundation models remain frozen.SourcesInflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Methods/Datasets (paragraph 1); Methods/Datasets (paragraph 2) |
| Context limits | A maximum input/context length for this exact evaluated configuration is not established by the inspected sources. · Not reported in inspected sourcesSources (2)Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing; Biodyn-AI/longevity-mechinterp README.md · Methods/Pipeline overview; Methods/Datasets; Methods/Models, baselines, and null calibrations; Methods/Interpretability blocks; Methods/Statistical safeguards; Methods/Exploratory methylation extension; Results/Randomized-weights ablations isolate the contribution of pretraining (asymmetric across models); inspected for explicit maximum input length (dataset lengths and family-wide limits are not substituted); README.md at pinned repository revision |
| Access | Official study implementation and usage documentation: https://github.com/Biodyn-AI/longevity-mechinterp/blob/5a61464632a3c3e8bebd396eb1ab17bce1dc2493/README.md. This pinned documentation revision is not automatically the evaluated weight revision.SourcesBiodyn-AI/longevity-mechinterp README.md · README.md; installation, model download and usage instructions |
| Code licence | No explicit code licence was established from the paper’s availability statement and inspected repository-root documentation. · Not reported in inspected sourcesSourcesBiodyn-AI/longevity-mechinterp README.md · README.md and repository-root licence-file search |
| Weights licence | The inspected model-access documentation does not explicitly identify terms for this exact evaluated checkpoint or fitted head; repository code terms are shown separately. · Not reported in inspected sourcesSourcesBiodyn-AI/longevity-mechinterp README.md · README.md; checkpoint/access documentation and licence scope |
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
20 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Diagram caption Conceptual input–method–output guide. Check the procedure text and linked evaluation for fitted components, additional inputs and exact settings. Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) Version: PMC archival version PMC13407579.1 | source checked automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram steps ["PBMC single-cell expression data with donor identities and age labels","Best frozen single-cell foundation model","Donor-age predictions and representation analyses"] Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) Version: PMC archival version PMC13407579.1 | source checked automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Diagram title Evaluated procedure (conceptual) Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) Version: PMC archival version PMC13407579.1 | source checked automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Model type Study-specific predictive method; this record is the paper-specific evaluated configuration. Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) Version: PMC archival version PMC13407579.1 | source checked automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Architecture / procedure Frozen scGPT and Geneformer representations feed age probes. The conventional baseline uses 50 principal components of gene expression with regression under the same donor-aware splits. Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) Version: PMC archival version PMC13407579.1 | source checked automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Weights licence The inspected model-access documentation does not explicitly identify terms for this exact evaluated checkpoint or fitted head; repository code terms are shown separately. Individual claims | Biodyn-AI/longevity-mechinterp README.md README.md; checkpoint/access documentation and licence scope Version: 5a61464632a3c3e8bebd396eb1ab17bce1dc2493 | unreported automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Biological inputs PBMC single-cell expression data with donor identities and age labels Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Methods/Statistical safeguards (paragraph 1); Methods/Datasets (paragraph 1) Version: PMC archival version PMC13407579.1 | source checked automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Outputs Donor-age predictions and representation analyses Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Discussion/Limitations (paragraph 1); Methods/Statistical safeguards (paragraph 1) Version: PMC archival version PMC13407579.1 | source checked automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Parameters This aggregate label does not identify a unique parameter count. The paper separately describes Geneformer V1-10M, V2-104M and V2-316M in its scaling analysis; those variants are not collapsed into one model identity. Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Discussion / Robustness and control analyses; cross-model summary label versus Cross-Geneformer-size benchmark Version: PMC archival version PMC13407579.1 | unreported automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Known versions / configuration Best frozen single-cell foundation model is the comparison-table label; that label does not specify an immutable weight revision. Individual claims | Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label. Version: PMC archival version PMC13407579.1 | unreported automated source review · 2026-09-16 Audit detailsPrimary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
Release 2026-09-17-d277315f7d76 · Record review: needs review
2 source records and release history
- Inflammation-linked aging signals in frozen single-cell foundation models: donor-aware detection and robustness testing · Original source · PMC archival version PMC13407579.1
- Biodyn-AI/longevity-mechinterp README.md · Original source · 5a61464632a3c3e8bebd396eb1ab17bce1dc2493
Technical metadata and extraction receipts
Stable ID: reported-model-148b613975b6eb
- areas
- cells-tissues
- entity level
- method
- version
- Not reported
- reported name
- Best frozen single-cell foundation model
- historical missing metadata
- version: not_reported_in_legacy_extract; checkpoint revision: not_reported_in_legacy_extract; training data: not_reported_in_legacy_extract; licence: not_reported_in_legacy_extract
- metadata review scope
- historical_missing_metadata preserves the original discovery state. Current descriptive evidence and missingness are recorded in profile.facts; numerical-result review is separate.
- legacy kinds
- model
- entity classification
- review date: 2026-09-17; rationale: This source-scoped entry preserves the method/configuration actually named in an evaluation. It is neither a global family identity nor proof of an immutable checkpoint; the linked evaluation retains adaptation, fitting and scoring details.; source ids: single-cell-aging-probes-2026; source locator: Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); Introduction/Why use frozen foundation models, given that simpler baselines predict equally well? (paragraph 1) | Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 2); Results/Foundation models detect age, but not better than a gene-expression PCA baseline (paragraph 1); ambiguities: Configuration means the source-labelled evaluated identity. It does not establish missing checkpoint hashes, default settings or equivalence to same-named records in other papers.