Models / rna-ml / rnafm / benchmarks

rna-ml/rnafm

Visibility: PUBLIC

Model card FilesBenchmarks

Group by

rna-secondary-structure-prediction rna-ml/rna-secondary-structure-prediction

RNA secondary structure (F1) rna-ml/rna-secondary-structure-f1

Drawn from: rna-secondary-structure-prediction

atlas-n100-v2

Metric	Value	95% CI	n	Rank	Best in catalogue	Source	Overlap	Board
Base-pair F1 (2-D pair head) ↑ higher is better · probe	0.580	[0.528, 0.631]	100	3 / 4	0.585	reproduced	overlap Train/test overlap declared by the evaluator: rna-ml/rnacentral	Board
Base-pair precision (2-D pair head) ↑ higher is better · probe · secondary	0.568	[0.514, 0.618]	100	—	0.591	reproduced	overlap Train/test overlap declared by the evaluator: rna-ml/rnacentral	Board
Base-pair recall (2-D pair head) ↑ higher is better · probe · secondary	0.633 best	[0.572, 0.691]	100	—	0.633	reproduced	overlap Train/test overlap declared by the evaluator: rna-ml/rnacentral	Board
Base-pair MCC (2-D pair head) ↑ higher is better · probe · secondary	0.588	[0.538, 0.639]	100	—	0.592	reproduced	overlap Train/test overlap declared by the evaluator: rna-ml/rnacentral	Board

Other Benchmarks drawn from no hub dataset you can see

Protein abundance from coding sequence (E. coli) rna-ml/coding-sequence-protein-abundance

atlas-n100-v1

Metric	Value	95% CI	n	Rank	Best in catalogue	Source	Overlap	Board
Spearman ρ (zero-shot pseudo-log-likelihood) ↑ higher is better · zero-shot	0.344	[0.146, 0.526]	100	4 / 11	0.552	reproduced	none declared	Board
Spearman ρ (ridge probe) ↑ higher is better · probe	0.543	[0.382, 0.669]	100	7 / 11	0.688	reproduced	none declared	Board
RMSE (ridge probe) ↓ lower is better · log10 ppm · probe	1.086	[0.958, 1.200]	100	7 / 11	0.973	reproduced	none declared	Board
Pearson r (ridge probe) ↑ higher is better · probe · secondary	0.548	[0.429, 0.648]	100	—	0.665	reproduced	none declared	Board
R² (ridge probe) ↑ higher is better · probe · secondary	0.266	[0.114, 0.385]	100	—	0.411	reproduced	none declared	Board