Model type
Long-context DNA model; this record is the paper-specific evaluated configuration.
Configuration
HyenaDNA
This HyenaDNA pipeline predicts polyadenylation sites from genomic sequence.
Sources
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Methods/Genome-wide discovery of de novo poly(A) sites using fine-tuned genome language models (paragraph 1); Methods (paragraph 2)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
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1)At a glance
Biological inputs
DNA windows around candidate polyadenylation/cleavage sites
Sources
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Results/Genome-wide prediction of novel poly(A) sites (paragraph 4); Results/Genome-wide prediction of novel poly(A) sites (paragraph 3)Outputs
Sequence-level or position-wise poly(A)-site predictions
Sources
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Results/Model interpretability: perturbation and attention analysis validate biological learning (paragraph 1); Abstract (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 |
|---|---|---|
| HyenaDNA: polyadenylation site detection few-shot Gene-Gene negative-set comparison Independent external evaluation · Evaluation metadata: needs review | ||
| 0.7510 AUC Unit: fraction · Direction: unknown | Uncertainty: not reported in legacy extract Scored: Not reported · Eligible: Not reported | source checkedPolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Table 1, Few-shot HyenaDNA row, G-G AUC column Source checking is not independent reproduction. |
How it works
How the evaluated method works
The study evaluates both few-shot classification and fine-tuning, and also builds a position-wise token classifier for extended gene regions.
Sources
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1)Underlying method and version boundaries
HyenaDNA models DNA at single-nucleotide resolution and publishes checkpoints with different context lengths. The family’s longest advertised context must not be assigned to a shorter-context checkpoint or to every downstream experiment.
Sources
HazyResearch/hyena-dna README.md · README.md; introduction, model description, pretrained-model and usage sections at pinned revisionWhat was evaluated
The linked evaluation record identifies HyenaDNA: polyadenylation site detection. Its dataset, split, adaptation and evidence origin remain attached to the reported results.
Sources
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · The named evaluation’s methods and comparison table; exact preserved evaluation IDs: evaluation-b2-polya-glm-2025Strengths and limitations
Strengths and considerations
- Signal-perturbation experiments examine recognition of canonical polyadenylation motifs and their location relative to cleavage sites.
Sources
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 2); Abstract (paragraph 1)
Limitations and conditions
- Few-shot and fine-tuned results are separate configurations; motif sensitivity alone is not experimental confirmation of a cleavage event.
Sources
PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Results/Genome-wide prediction of novel poly(A) sites (paragraph 3); Methods/Validating biological feature learning through interpretability analysis (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-953007693fb72aSpecifications
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 | Long-context DNA model; this record is the paper-specific evaluated configuration.SourcesHazyResearch/hyena-dna README.md · README.md model description |
| Architecture / procedure | The study evaluates both few-shot classification and fine-tuning, and also builds a position-wise token classifier for extended gene regions.SourcesPolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1) |
| Biological inputs | DNA windows around candidate polyadenylation/cleavage sitesSourcesPolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Results/Genome-wide prediction of novel poly(A) sites (paragraph 4); Results/Genome-wide prediction of novel poly(A) sites (paragraph 3) |
| Outputs | Sequence-level or position-wise poly(A)-site predictionsSourcesPolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Results/Model interpretability: perturbation and attention analysis validate biological learning (paragraph 1); Abstract (paragraph 1) |
| Parameters | An aggregate parameter total for this exact evaluated configuration is not established by the inspected sources. · Not reported in inspected sourcesSources (2)PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models; HazyResearch/hyena-dna README.md · Methods; Methods/An end-to-end framework for poly(A) site classification with genome language models; Methods/Genome-wide discovery of de novo poly(A) sites using fine-tuned genome language models; Methods/Fine-grained poly(A) site identification through token-level modeling; Methods/Validating biological feature learning through interpretability analysis; inspected for aggregate parameter count (component sizes are not added without an exact configuration); README.md at pinned repository revision |
| Known versions / configuration | HyenaDNA is the comparison-table label; that label does not specify an immutable weight revision. · Not reported in inspected sourcesSourcesPolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label. |
| Training data / fitting | GENCODE v38 positive poly(A) annotations and motif-matched negative sequences form balanced Gene-Gene and Gene-Intergene datasets, each with 50,182 positive and 50,182 negative examples. Five folds use 60% training,20% validation and 20% test; few-shot prototypes and full fine-tuning are distinct settings.SourcesPolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Methods/An end-to-end framework for poly(A) site classification with genome language models (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 2) |
| Context limits | 101-bp windows, with position 51 explicitly set to adenine in both positive and negative examples.SourcesPolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Methods/An end-to-end framework for poly(A) site classification with genome language models (paragraph 1); Methods/An end-to-end framework for poly(A) site classification with genome language models (paragraph 2) |
| Access | Official upstream implementation and usage documentation: https://github.com/HazyResearch/hyena-dna/blob/d553021b483b82980aa4b868b37ec2d4332e198a/README.md. This pinned documentation revision is not automatically the evaluated weight revision.SourcesHazyResearch/hyena-dna README.md · README.md; installation, model download and usage instructions |
| Code licence | Apache 2.0 (upstream repository code at the cited revision; this does not establish every dependency or historical checkpoint licence).SourcesHazyResearch/hyena-dna 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 sourcesSourcesHazyResearch/hyena-dna 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.
21 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 | PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1) Version: journal full text 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 ["DNA windows around candidate polyadenylation/cleavage sites","HyenaDNA","Sequence-level or position-wise poly(A)-site predictions"] Individual claims | PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1) Version: journal full text 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 | PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1) Version: journal full text 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 Long-context DNA model; this record is the paper-specific evaluated configuration. Individual claims | HazyResearch/hyena-dna README.md README.md model description Version: d553021b483b82980aa4b868b37ec2d4332e198a | 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 The study evaluates both few-shot classification and fine-tuning, and also builds a position-wise token classifier for extended gene regions. Individual claims | PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1) Version: journal full text 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 | HazyResearch/hyena-dna README.md README.md; checkpoint/access documentation and licence scope Version: d553021b483b82980aa4b868b37ec2d4332e198a | 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 windows around candidate polyadenylation/cleavage sites Individual claims | PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models Results/Genome-wide prediction of novel poly(A) sites (paragraph 4); Results/Genome-wide prediction of novel poly(A) sites (paragraph 3) Version: journal full text 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-level or position-wise poly(A)-site predictions Individual claims | PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models Results/Model interpretability: perturbation and attention analysis validate biological learning (paragraph 1); Abstract (paragraph 1) Version: journal full text 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 An aggregate parameter total for this exact evaluated configuration is not established by the inspected sources. Individual claims | HazyResearch/hyena-dna README.md Methods; Methods/An end-to-end framework for poly(A) site classification with genome language models; Methods/Genome-wide discovery of de novo poly(A) sites using fine-tuned genome language models; Methods/Fine-grained poly(A) site identification through token-level modeling; Methods/Validating biological feature learning through interpretability analysis; inspected for aggregate parameter count (component sizes are not added without an exact configuration); README.md at pinned repository revision Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: d553021b483b82980aa4b868b37ec2d4332e198a | 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 |
| Parameters An aggregate parameter total for this exact evaluated configuration is not established by the inspected sources. Individual claims | PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models Methods; Methods/An end-to-end framework for poly(A) site classification with genome language models; Methods/Genome-wide discovery of de novo poly(A) sites using fine-tuned genome language models; Methods/Fine-grained poly(A) site identification through token-level modeling; Methods/Validating biological feature learning through interpretability analysis; inspected for aggregate parameter count (component sizes are not added without an exact configuration); README.md at pinned repository revision Shared locator for this statement’s cited sources; not a separate locator for each citation. Version: journal full text 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
Release 2026-09-17-d277315f7d76 · Record review: needs review
3 source records and release history
- PolyA-GLM: A comprehensive framework for De novo polyadenylation site prediction using genome language models · Original source · journal full text in PMC
- HazyResearch/hyena-dna README.md · Original source · d553021b483b82980aa4b868b37ec2d4332e198a
- HazyResearch/hyena-dna LICENSE · Original source · d553021b483b82980aa4b868b37ec2d4332e198a
Technical metadata and extraction receipts
Stable ID: reported-model-953007693fb72a
- areas
- dna-genomes
- entity level
- method
- version
- not stated in table
- reported name
- HyenaDNA
- 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: polya-glm-2025; evidence-reported-base-hyenadna-readme-md; source locator: Abstract (paragraph 1); Results/Model performance: few-shot capabilities and fine-tuning improvements (paragraph 1) | README.md model description | Methods/Genome-wide discovery of de novo poly(A) sites using fine-tuned genome language models (paragraph 1); Methods (paragraph 2); 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.