Configuration

Splice-MS1024

Published RNA language model backbone, fine-tuned per task by the BEACON authors under their protocol.

Overview

Published RNA language model backbone, fine-tuned per task by the BEACON authors under their protocol.

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

Evaluations and results

Release 2026-09-17-1a18ca2c038a · 13 evaluations · 13 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
Splice-MS1024 on BEACON APA: Alternative polyadenylation isoform prediction Configuration: Splice-MS1024Task: BEACON APA: Alternative polyadenylation isoform predictionDataset subset: APARENT (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 145,463/33,170/49,755; dataset APARENT. Author-reported evaluation · Evaluation metadata: source checked
60.03(3.42)% r2 Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 3.42 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(APA) Source checking is not independent reproduction.
Splice-MS1024 on BEACON CMP: Contact map prediction Configuration: Splice-MS1024Task: BEACON CMP: Contact map predictionDataset subset: RNAcontact (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 188/23/80; dataset RNAcontact. Author-reported evaluation · Evaluation metadata: source checked
47.32(3.16)% precision_at_l Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 3.16 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(CMP) Source checking is not independent reproduction.
Splice-MS1024 on BEACON CRI-Off: CRISPR off-target effect prediction Configuration: Splice-MS1024Task: BEACON CRI-Off: CRISPR off-target effect predictionDataset subset: DeepCRISPR (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 14,223/2,032/4,064; dataset DeepCRISPR. Author-reported evaluation · Evaluation metadata: source checked
5.00(0.71)% spearman_corr Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 0.71 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(CRI-Off) Source checking is not independent reproduction.
Splice-MS1024 on BEACON CRI-On: CRISPR on-target efficiency prediction Configuration: Splice-MS1024Task: BEACON CRI-On: CRISPR on-target efficiency predictionDataset subset: DeepCRISPR (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 1,453/207/416; dataset DeepCRISPR. Author-reported evaluation · Evaluation metadata: source checked
27.59(4.61)% spearman_corr Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 4.61 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(CRI-On) Source checking is not independent reproduction.
Splice-MS1024 on BEACON DMP: Distance map prediction Configuration: Splice-MS1024Task: BEACON DMP: Distance map predictionDataset subset: RNAcontact (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 188/23/80; dataset RNAcontact. Author-reported evaluation · Evaluation metadata: source checked
55.89(0.48)% r2 Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 0.48 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(DMP) Source checking is not independent reproduction.
Splice-MS1024 on BEACON Modif: RNA modification site prediction Configuration: Splice-MS1024Task: BEACON Modif: RNA modification site predictionDataset subset: MultiRM (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 304,661/3,599/1,200; dataset MultiRM. Author-reported evaluation · Evaluation metadata: source checked
53.45(6.25)% auc Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 6.25 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(Modif) Source checking is not independent reproduction.
Splice-MS1024 on BEACON MRL: Mean ribosome loading prediction Configuration: Splice-MS1024Task: BEACON MRL: Mean ribosome loading predictionDataset subset: Optimus (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 76,319/7,600/7,600; dataset Optimus. Author-reported evaluation · Evaluation metadata: source checked
67.15(30.54)% r2 Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 30.54 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(MRL) Source checking is not independent reproduction.
Splice-MS1024 on BEACON ncRNA: Non-coding RNA family classification Configuration: Splice-MS1024Task: BEACON ncRNA: Non-coding RNA family classificationDataset subset: Noorul's ncRNA set (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 5,679/650/2,400; dataset Noorul's ncRNA set. Author-reported evaluation · Evaluation metadata: source checked
96.05(0.777)% accuracy Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 0.777 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(ncRNA) Source checking is not independent reproduction.
Splice-MS1024 on BEACON PRS: Programmable RNA switch prediction Configuration: Splice-MS1024Task: BEACON PRS: Programmable RNA switch predictionDataset subset: Angenent-Mari's switch set (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 73,227/9,153/9,154; dataset Angenent-Mari's switch set. Author-reported evaluation · Evaluation metadata: source checked
57.72(0.45)% r2 Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 0.45 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(PRS) Source checking is not independent reproduction.
Splice-MS1024 on BEACON SPL: Splice site prediction Configuration: Splice-MS1024Task: BEACON SPL: Splice site predictionDataset subset: SpliceAI (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 144,628/18,078/16,505; dataset SpliceAI. Author-reported evaluation · Evaluation metadata: source checked
48.52(0.49)% top_k_accuracy Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 0.49 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(SPL) Source checking is not independent reproduction.
Splice-MS1024 on BEACON SSI: Structure score imputation Configuration: Splice-MS1024Task: BEACON SSI: Structure score imputationDataset subset: StructureImpute (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 14,049/1,756/3,095; dataset StructureImpute. Author-reported evaluation · Evaluation metadata: source checked
39.22(0.02)% r2 Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 0.02 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(SSI) Source checking is not independent reproduction.
Splice-MS1024 on BEACON SSP: Secondary structure prediction Configuration: Splice-MS1024Task: BEACON SSP: Secondary structure predictionDataset subset: bpRNA-1m (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 10,814/1,300/1,305; dataset bpRNA-1m. Author-reported evaluation · Evaluation metadata: source checked
68.26(0.20)% f1 Unit: percent · Direction: higher	Uncertainty: type: standard deviation; value: 0.20 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(SSP) Source checking is not independent reproduction.
Splice-MS1024 on BEACON VDP: Vaccine degradation prediction Configuration: Splice-MS1024Task: BEACON VDP: Vaccine degradation predictionDataset subset: OpenVaccine (BEACON split) BEACON harness, fixed downstream head per task. Train/validation/test 2,155/245/629; dataset OpenVaccine. Author-reported evaluation · Evaluation metadata: source checked
0.313(0.000) mcrmse Unit: error · Direction: lower	Uncertainty: type: standard deviation; value: 0.000 Scored: Not reported · Eligible: Not reported	source checkedBEACON: Benchmark for Comprehensive RNA Tasks and Language Models (arXiv:2406.10391v2) · Table3,p.8,row(Splice-MS1024),column(VDP) 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