Method

Naive Baseline

Baseline trained from scratch by the mRNABench authors.

Overview

Baseline trained from scratch by the mRNABench authors.

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

Evaluations and results

Release 2026-09-17-134cd1815de8 · 10 evaluations · 10 metric rows. Different protocols are not a single leaderboard. Where several source tables report the same metric, the published comparisons above offer a pooled view that names what it does not hold constant.

Evaluation setupMetricEvidence origin

Results grouped by the exact reported evaluation
Metric and finding	Coverage and uncertainty	Evidence
Naive Baseline on mRNABench ECLIP: eCLIP binding site prediction Method: Naive BaselineTask: mRNABench ECLIP: eCLIP binding site predictionDataset subset: mRNABench eCLIP (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
43.4% auprc Unit: percent · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(eCLIP) Source checking is not independent reproduction.
Naive Baseline on mRNABench GO: Gene Ontology term prediction Method: Naive BaselineTask: mRNABench GO: Gene Ontology term predictionDataset subset: mRNABench GO (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
23.4% auprc Unit: percent · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(GO) Source checking is not independent reproduction.
Naive Baseline on mRNABench HL: mRNA half life Method: Naive BaselineTask: mRNABench HL: mRNA half lifeDataset subset: mRNABench HL (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
0.20 pearson_r Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(HL) Source checking is not independent reproduction.
Naive Baseline on mRNABench MRL-HL-PAIR: Paired mean ribosome load and half life Method: Naive BaselineTask: mRNABench MRL-HL-PAIR: Paired mean ribosome load and half lifeDataset subset: mRNABench MRL-HL-Pair (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
0.51 pearson_r Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(MRL-HL-Pair) Source checking is not independent reproduction.
Naive Baseline on mRNABench MRL-MPRA: Mean ribosome load on an MPRA library Method: Naive BaselineTask: mRNABench MRL-MPRA: Mean ribosome load on an MPRA libraryDataset subset: mRNABench MRL MPRA (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
0.64 pearson_r Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(MRL MPRA) Source checking is not independent reproduction.
Naive Baseline on mRNABench MRL: Mean ribosome load Method: Naive BaselineTask: mRNABench MRL: Mean ribosome loadDataset subset: mRNABench MRL (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
0.14 pearson_r Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(MRL) Source checking is not independent reproduction.
Naive Baseline on mRNABench MRNA-LOC-LR: mRNA localisation, long range Method: Naive BaselineTask: mRNABench MRNA-LOC-LR: mRNA localisation, long rangeDataset subset: mRNABench mRNA Loc-LR (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
62.2% auprc Unit: percent · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(mRNA Loc-LR) Source checking is not independent reproduction.
Naive Baseline on mRNABench MRNA-LOC-SR: mRNA localisation, short range Method: Naive BaselineTask: mRNABench MRNA-LOC-SR: mRNA localisation, short rangeDataset subset: mRNABench mRNA Loc-SR (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
60.1% auprc Unit: percent · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(mRNA Loc-SR) Source checking is not independent reproduction.
Naive Baseline on mRNABench PROT-LOC: Protein localisation Method: Naive BaselineTask: mRNABench PROT-LOC: Protein localisationDataset subset: mRNABench Prot Loc (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
21.5% auprc Unit: percent · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(Prot Loc) Source checking is not independent reproduction.
Naive Baseline on mRNABench VEP: Variant effect prediction Method: Naive BaselineTask: mRNABench VEP: Variant effect predictionDataset subset: mRNABench VEP (mRNABench split) A linear probe over frozen embeddings, scored as the mean over ten random seeds. Author-reported evaluation · Evaluation metadata: source checked
19.5% auprc Unit: percent · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedmRNABench: A curated benchmark for mature mRNA property and function prediction · Table 2, row(Naive Baseline), column(VEP) Source checking is not 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