Configuration

NTv3 (full fine-tuning, GFMBench-API)

NTv3, a configuration of Nucleotide Transformer v3: Nucleotide Transformer v3, about 8M parameters; Hugging Face checkpoint; full fine-tuning; full fine-tuning for 3 epochs.

10 evaluations · 40 results

Overview

An explanatory profile has not yet been reviewed for this record.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

Evaluations and results

10 evaluations · 40 results. Different protocols are not a single leaderboard.

Sorted by Accuracy (test set) (higher is better). The best value in each column is highlighted. Decimals are rounded for display; each value links to the printed value and its source.

Tested configuration	Protocol and dataset	Accuracy (test set)	AUPRC (test set)	AUROC (test set)	MCC (test set)
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API Var Bench Non Coding Pathogenicity, supervised, Var Bench Non Coding Pathogenicity test set (GFMBench-API)	0.981 (best in column)	0.931	0.952	0.894 (best in column)
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API GUE Promoter All, supervised, GUE Promoter All test set (GFMBench-API)	0.94	0.988 (best in column)	0.987 (best in column)	0.879
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API GUE Splice Site, supervised, GUE Splice Site test set (GFMBench-API)	0.932	0.968	0.981	0.884
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API GUE Transcription Factor, supervised, GUE Transcription Factor test set (GFMBench-API)	0.808	0.884	0.89	0.617
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API Var Bench Expression, supervised, Var Bench Expression test set (GFMBench-API)	0.674	0.706	0.742	0.349
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API LRB Variant Effect Causal eQTL, supervised, LRB Variant Effect Causal eQTL test set (GFMBench-API)	0.625	0.702	0.68	0.28
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API Var Bench Coding Pathogenicity, supervised, Var Bench Coding Pathogenicity test set (GFMBench-API)	0.622	0.579	0.683	0.204
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API Var Bench Common Vs Rare, supervised, Var Bench Common Vs Rare test set (GFMBench-API)	0.588	0.636	0.634	0.18
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API Var Bench meQTL, supervised, Var Bench meQTL test set (GFMBench-API)	0.573	0.578	0.656	0.118
NTv3 (full fine-tuning, GFMBench-API)Evaluation	GFMBench-API Var Bench sQTL, supervised, Var Bench sQTL test set (GFMBench-API)	0.502	0.546	0.54	0.0402

All 40 result rows with coverage, uncertainty and sources

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Promoter All, supervised Dataset: GUE Promoter All test set (GFMBench-API)	0.94 Accuracy (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Promoter All (supervised) lit-genomics-20261011-gfmbench-protocol-gue-promoter-all-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Promoter All', row 'NTv3', column 'Accuracy'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Promoter All, supervised Dataset: GUE Promoter All test set (GFMBench-API)	0.988 AUPRC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Promoter All (supervised) lit-genomics-20261011-gfmbench-protocol-gue-promoter-all-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Promoter All', row 'NTv3', column 'AUPRC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Promoter All, supervised Dataset: GUE Promoter All test set (GFMBench-API)	0.987 AUROC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Promoter All (supervised) lit-genomics-20261011-gfmbench-protocol-gue-promoter-all-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Promoter All', row 'NTv3', column 'AUROC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Promoter All, supervised Dataset: GUE Promoter All test set (GFMBench-API)	0.879 MCC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Promoter All (supervised) lit-genomics-20261011-gfmbench-protocol-gue-promoter-all-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Promoter All', row 'NTv3', column 'MCC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Splice Site, supervised Dataset: GUE Splice Site test set (GFMBench-API)	0.932 Accuracy (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Splice Site (supervised) lit-genomics-20261011-gfmbench-protocol-gue-splice-site-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Splice Site', row 'NTv3', column 'Accuracy'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Splice Site, supervised Dataset: GUE Splice Site test set (GFMBench-API)	0.968 AUPRC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Splice Site (supervised) lit-genomics-20261011-gfmbench-protocol-gue-splice-site-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Splice Site', row 'NTv3', column 'AUPRC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Splice Site, supervised Dataset: GUE Splice Site test set (GFMBench-API)	0.981 AUROC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Splice Site (supervised) lit-genomics-20261011-gfmbench-protocol-gue-splice-site-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Splice Site', row 'NTv3', column 'AUROC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Splice Site, supervised Dataset: GUE Splice Site test set (GFMBench-API)	0.884 MCC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Splice Site (supervised) lit-genomics-20261011-gfmbench-protocol-gue-splice-site-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Splice Site', row 'NTv3', column 'MCC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Transcription Factor, supervised Dataset: GUE Transcription Factor test set (GFMBench-API)	0.808 Accuracy (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Transcription Factor (supervised) lit-genomics-20261011-gfmbench-protocol-gue-transcription-factor-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Transcription Factor', row 'NTv3', column 'Accuracy'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Transcription Factor, supervised Dataset: GUE Transcription Factor test set (GFMBench-API)	0.884 AUPRC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Transcription Factor (supervised) lit-genomics-20261011-gfmbench-protocol-gue-transcription-factor-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Transcription Factor', row 'NTv3', column 'AUPRC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Transcription Factor, supervised Dataset: GUE Transcription Factor test set (GFMBench-API)	0.89 AUROC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Transcription Factor (supervised) lit-genomics-20261011-gfmbench-protocol-gue-transcription-factor-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Transcription Factor', row 'NTv3', column 'AUROC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API GUE Transcription Factor, supervised Dataset: GUE Transcription Factor test set (GFMBench-API)	0.617 MCC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API GUE Transcription Factor (supervised) lit-genomics-20261011-gfmbench-protocol-gue-transcription-factor-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'GUE Transcription Factor', row 'NTv3', column 'MCC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API LRB Variant Effect Causal eQTL, supervised Dataset: LRB Variant Effect Causal eQTL test set (GFMBench-API)	0.625 Accuracy (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API LRB Variant Effect Causal eQTL (supervised) lit-genomics-20261011-gfmbench-protocol-lrb-variant-effect-causal-eqtl-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'LRB Variant Effect Causal eQTL', row 'NTv3', column 'Accuracy'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API LRB Variant Effect Causal eQTL, supervised Dataset: LRB Variant Effect Causal eQTL test set (GFMBench-API)	0.702 AUPRC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API LRB Variant Effect Causal eQTL (supervised) lit-genomics-20261011-gfmbench-protocol-lrb-variant-effect-causal-eqtl-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'LRB Variant Effect Causal eQTL', row 'NTv3', column 'AUPRC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API LRB Variant Effect Causal eQTL, supervised Dataset: LRB Variant Effect Causal eQTL test set (GFMBench-API)	0.68 AUROC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API LRB Variant Effect Causal eQTL (supervised) lit-genomics-20261011-gfmbench-protocol-lrb-variant-effect-causal-eqtl-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'LRB Variant Effect Causal eQTL', row 'NTv3', column 'AUROC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API LRB Variant Effect Causal eQTL, supervised Dataset: LRB Variant Effect Causal eQTL test set (GFMBench-API)	0.28 MCC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API LRB Variant Effect Causal eQTL (supervised) lit-genomics-20261011-gfmbench-protocol-lrb-variant-effect-causal-eqtl-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'LRB Variant Effect Causal eQTL', row 'NTv3', column 'MCC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Coding Pathogenicity, supervised Dataset: Var Bench Coding Pathogenicity test set (GFMBench-API)	0.622 Accuracy (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Coding Pathogenicity (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-coding-pathogenicity-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Coding Pathogenicity', row 'NTv3', column 'Accuracy'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Coding Pathogenicity, supervised Dataset: Var Bench Coding Pathogenicity test set (GFMBench-API)	0.579 AUPRC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Coding Pathogenicity (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-coding-pathogenicity-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Coding Pathogenicity', row 'NTv3', column 'AUPRC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Coding Pathogenicity, supervised Dataset: Var Bench Coding Pathogenicity test set (GFMBench-API)	0.683 AUROC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Coding Pathogenicity (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-coding-pathogenicity-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Coding Pathogenicity', row 'NTv3', column 'AUROC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Coding Pathogenicity, supervised Dataset: Var Bench Coding Pathogenicity test set (GFMBench-API)	0.204 MCC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Coding Pathogenicity (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-coding-pathogenicity-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Coding Pathogenicity', row 'NTv3', column 'MCC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Common Vs Rare, supervised Dataset: Var Bench Common Vs Rare test set (GFMBench-API)	0.588 Accuracy (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Common Vs Rare (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-common-vs-rare-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Common Vs Rare', row 'NTv3', column 'Accuracy'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Common Vs Rare, supervised Dataset: Var Bench Common Vs Rare test set (GFMBench-API)	0.636 AUPRC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Common Vs Rare (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-common-vs-rare-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Common Vs Rare', row 'NTv3', column 'AUPRC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Common Vs Rare, supervised Dataset: Var Bench Common Vs Rare test set (GFMBench-API)	0.634 AUROC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Common Vs Rare (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-common-vs-rare-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Common Vs Rare', row 'NTv3', column 'AUROC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Common Vs Rare, supervised Dataset: Var Bench Common Vs Rare test set (GFMBench-API)	0.18 MCC (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Common Vs Rare (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-common-vs-rare-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Common Vs Rare', row 'NTv3', column 'MCC'
Configuration: NTv3 (full fine-tuning, GFMBench-API)	Protocol: GFMBench-API Var Bench Expression, supervised Dataset: Var Bench Expression test set (GFMBench-API)	0.674 Accuracy (test set) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source NTv3 on GFMBench-API Var Bench Expression (supervised) lit-genomics-20261011-gfmbench-protocol-var-bench-expression-supervised Aggregation: Not reported GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Table 1, task 'Var Bench Expression', row 'NTv3', column 'Accuracy'

Source checking is not independent reproduction. Release 2026-10-22-1b7cf027e4f6.

Use this configuration

How it works, versions and access

Strengths, limitations and unresolved questions

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.

0 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-10-22-1b7cf027e4f6
Property and statement	Original source and location	Review and provenance

No evidence rows match these filters. Choose another scope or clear the search.

How to read the evidence

Sources and history

Release 2026-10-22-1b7cf027e4f6 · Record review: source checked

1 source record and release history

GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models · Original source · bioRxiv preprint version 2, posted 2026-08-11

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: lit-genomics-20261011-gfmbench-config-ntv3-full-fine-tuning

areas: dna-genomes
contexts: research
method types: foundation_model
reported name: NTv3
foundation model eligible: true
parameters: Nucleotide Transformer v3, about 8M parameters; Hugging Face checkpoint; full fine-tuning; full fine-tuning for 3 epochs
access reuse: Public checkpoint (Hugging Face; Evo2 via NVIDIA BioNeMo)
source locator: Benchmark Results sections A and B
missing metadata: version: reason: unreported; note: Checkpoint revision not printed; parameter count only
provenance note: NTv3 as evaluated in GFMBench-API, full fine-tuning.

Related records

configuration of: Nucleotide Transformer v3
system: NTv3 on GFMBench-API GUE Promoter All (supervised)
system: NTv3 on GFMBench-API GUE Splice Site (supervised)
system: NTv3 on GFMBench-API GUE Transcription Factor (supervised)
system: NTv3 on GFMBench-API LRB Variant Effect Causal eQTL (supervised)
system: NTv3 on GFMBench-API Var Bench Coding Pathogenicity (supervised)
system: NTv3 on GFMBench-API Var Bench Common Vs Rare (supervised)
system: NTv3 on GFMBench-API Var Bench Expression (supervised)
system: NTv3 on GFMBench-API Var Bench meQTL (supervised)
system: NTv3 on GFMBench-API Var Bench Non Coding Pathogenicity (supervised)
system: NTv3 on GFMBench-API Var Bench sQTL (supervised)

Suggest a correction