Model type
Study-specific predictive method; this record is the paper-specific evaluated configuration.
Configuration
PhyloGPN
PhyloGPN learns genomic representations with a phylogeny-aware objective and predicts from single sequences.
Sources
A Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Discussion and Conclusion (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
A Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Discussion and Conclusion (paragraph 1)Overview
Biological inputs
DNA sequence at inference; multispecies alignment and tree information during training
Sources
A Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Background and Related Work/Molecular Phylogenetics (paragraph 1)Outputs
Sequence representations and variant-effect scores
Sources
A Phylogenetic Approach to Genomic Language Modeling · Results/Embedding Evaluation (paragraph 1); Introduction (paragraph 3)Access
Official study implementation and usage documentation: https://github.com/songlab-cal/gpn/blob/6f28c81bcbfe7d65cb6d8ece9ce88f87ca583791/README.md. This pinned documentation revision is not automatically the evaluated weight revision.
Sources
songlab-cal/gpn README.md · README.md; installation, model download and usage instructionslimited source coverage · Automated source review, 2026-09-16. All specifications and missing details
Evaluations and results
1 evaluation · 1 result. Different protocols are not a single leaderboard.
Filter evaluations
Applied filters: All linked evaluations
| Tested configuration | Protocol and dataset | Finding | Evidence and details |
|---|---|---|---|
| Configuration: PhyloGPN | Task: ClinVar 3-prime UTR variant classification Dataset: ClinVar 3-prime UTR variants | 0.94 AUROC fraction · unknown Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible | Author-reported evaluation · Source checkedMethods, coverage and sourcePhyloGPN: ClinVar 3-prime UTR variant classification log-likelihood-ratio scoring Aggregation: Not reported A Phylogenetic Approach to Genomic Language Modeling · Table 1, 3-prime UTR row, PhyloGPN AUROC column |
Source checking is not independent reproduction. Release 2026-09-29-06401fd5b220.
Use this model
How it works, versions and access
How it works
How the evaluated method works
Multispecies whole-genome alignments and phylogenetic trees enter the training loss to model nucleotide evolution. They are not required as inputs for downstream single-sequence prediction.
Sources
A Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Discussion and Conclusion (paragraph 1)What was evaluated
The linked evaluation record identifies PhyloGPN: ClinVar 3-prime UTR variant classification. Its dataset, split, adaptation and evidence origin remain attached to the reported results.
Sources
A Phylogenetic Approach to Genomic Language Modeling · The named evaluation’s methods and comparison table; exact preserved evaluation IDs: evaluation-b2-phylogpn-2025Strengths, limitations and unresolved questions
Strengths and limitations
Strengths and considerations
- Uses evolutionary information during training while retaining sequence-only inference.
Sources
A Phylogenetic Approach to Genomic Language Modeling · Background and Related Work/Transfer Learning and Genomic Language Models (gLMs) (paragraph 2); Results/Embedding Evaluation (paragraph 4)
Limitations and conditions
- Sequence-only inference does not imply sequence-only training; alignment quality and phylogenetic assumptions remain relevant.
Sources
A Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Background and Related Work/Molecular Phylogenetics (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-cbbe04b826ceffSpecifications
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.SourcesA Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Discussion and Conclusion (paragraph 1) |
| Architecture / procedure | Multispecies whole-genome alignments and phylogenetic trees enter the training loss to model nucleotide evolution. They are not required as inputs for downstream single-sequence prediction.SourcesA Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Discussion and Conclusion (paragraph 1) |
| Biological inputs | DNA sequence at inference; multispecies alignment and tree information during trainingSourcesA Phylogenetic Approach to Genomic Language Modeling · Abstract (paragraph 1); Background and Related Work/Molecular Phylogenetics (paragraph 1) |
| Outputs | Sequence representations and variant-effect scoresSourcesA Phylogenetic Approach to Genomic Language Modeling · Results/Embedding Evaluation (paragraph 1); Introduction (paragraph 3) |
| Parameters | About 83M stored parameters; reverse-complement weight tying means the number of free parameters is roughly half the weights and biases, plus layer-normalisation parameters.SourcesA Phylogenetic Approach to Genomic Language Modeling · Methods/Architecture and Training (paragraph 1); Methods/Loss Function (paragraph 1) |
| Known versions / configuration | PhyloGPN is the comparison-table label; that label does not specify an immutable weight revision. · Not reported in inspected sourcesSourcesA Phylogenetic Approach to Genomic Language Modeling · Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label. |
| Training data / fitting | Multispecies whole-genome alignment with explicit phylogenetic modelling.SourcesA Phylogenetic Approach to Genomic Language Modeling · Methods/Training Data (paragraph 2); Methods/Training Data (paragraph 1) |
| Context limits | A 481-bp receptive field from 40 dilated residual blocks; training consumes approximately 10-kb alignment blocks, a different quantity.SourcesA Phylogenetic Approach to Genomic Language Modeling · Methods/Architecture and Training (paragraph 1); Methods/Architecture and Training (paragraph 2) |
| Access | Official study implementation and usage documentation: https://github.com/songlab-cal/gpn/blob/6f28c81bcbfe7d65cb6d8ece9ce88f87ca583791/README.md. This pinned documentation revision is not automatically the evaluated weight revision.Sourcessonglab-cal/gpn README.md · README.md; installation, model download and usage instructions |
| Code licence | MIT (study repository code at the cited revision; this does not establish every dependency or historical checkpoint licence).Sourcessonglab-cal/gpn LICENSE · LICENSE; complete licence text |
| 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 sourcesSourcessonglab-cal/gpn README.md · README.md; checkpoint/access documentation and licence scope |
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
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.
19 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 | A Phylogenetic Approach to Genomic Language Modeling Abstract (paragraph 1); Discussion and Conclusion (paragraph 1) Version: preprint version in PMC | 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
| A Phylogenetic Approach to Genomic Language Modeling Abstract (paragraph 1); Discussion and Conclusion (paragraph 1) Version: preprint version in PMC | 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 | A Phylogenetic Approach to Genomic Language Modeling Abstract (paragraph 1); Discussion and Conclusion (paragraph 1) Version: preprint version in PMC | 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 | A Phylogenetic Approach to Genomic Language Modeling Abstract (paragraph 1); Discussion and Conclusion (paragraph 1) Version: preprint version in PMC | 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 Multispecies whole-genome alignments and phylogenetic trees enter the training loss to model nucleotide evolution. They are not required as inputs for downstream single-sequence prediction. Individual claims | A Phylogenetic Approach to Genomic Language Modeling Abstract (paragraph 1); Discussion and Conclusion (paragraph 1) Version: preprint version in PMC | 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 | songlab-cal/gpn README.md README.md; checkpoint/access documentation and licence scope Version: 6f28c81bcbfe7d65cb6d8ece9ce88f87ca583791 | 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 DNA sequence at inference; multispecies alignment and tree information during training Individual claims | A Phylogenetic Approach to Genomic Language Modeling Abstract (paragraph 1); Background and Related Work/Molecular Phylogenetics (paragraph 1) Version: preprint version in PMC | 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 Sequence representations and variant-effect scores Individual claims | A Phylogenetic Approach to Genomic Language Modeling Results/Embedding Evaluation (paragraph 1); Introduction (paragraph 3) Version: preprint version in PMC | 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 About 83M stored parameters; reverse-complement weight tying means the number of free parameters is roughly half the weights and biases, plus layer-normalisation parameters. Individual claims | A Phylogenetic Approach to Genomic Language Modeling Methods/Architecture and Training (paragraph 1); Methods/Loss Function (paragraph 1) Version: preprint version in PMC | 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 |
| Known versions / configuration PhyloGPN is the comparison-table label; that label does not specify an immutable weight revision. Individual claims | A Phylogenetic Approach to Genomic Language Modeling Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label. Version: preprint version in PMC | 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
View linked audit checks and correction history
Release 2026-09-29-06401fd5b220 · Record review: needs review
3 source records and release history
- A Phylogenetic Approach to Genomic Language Modeling · Original source · preprint version in PMC
- songlab-cal/gpn README.md · Original source · 6f28c81bcbfe7d65cb6d8ece9ce88f87ca583791
- songlab-cal/gpn LICENSE · Original source · 6f28c81bcbfe7d65cb6d8ece9ce88f87ca583791
Technical metadata and extraction receipts
Stable ID: reported-model-cbbe04b826ceff
- areas
- dna-genomes
- entity level
- method
- version
- not stated in table
- reported name
- PhyloGPN
- historical missing metadata
- 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: phylogpn-2025; source locator: Abstract (paragraph 1); Discussion and Conclusion (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.