Contested result

Geneformer and scGPT: Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13) reverses the order

Medium severity
Not yet reviewed
Ordering reversal
Different tasks across studies

Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13) ranks Geneformer above scGPT in 3 of 4 comparisons; other studies rank scGPT above Geneformer in 95 of 98.

An automatic lead, not a finding of error. The results below are shown as published, everywhere on the site, until a reviewer decides otherwise.

Which studies agree

3 papers rank scGPT above Geneformer in 95 of their 100 comparisons and Geneformer higher in 3; 2 are ties. Of those papers, Deep-learning-based gene perturbation… ranks Geneformer higher on its own (2 of 2). Wu et al. 2025 ranks Geneformer above scGPT in 3 of its 4 comparisons and scGPT higher in 1.

Rank scGPT higher (2): Kedzierska et al. 2025, Parameter-Efficient Fine-Tuning…
Rank Geneformer higher (2): Wu et al. 2025 (disagrees), Deep-learning-based gene perturbation…
Mixed (0): none

Each paper sits where at least two thirds of its comparisons point; Mixed means neither model reaches two thirds. The larger, outlined dot is the paper that disagrees.

scGPT higherTieGeneformer higherNumbers are counts of comparisons; the outlined bar is the paper that disagrees.

Papers compared (4)
Paper	Role	Ranks scGPT higher	Ranks Geneformer higher	Ties
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13) · DOI	Disagrees	1	3	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx) and 1 more record · DOI	Agrees	94	1	2
Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines and 4 more records · DOI	Agrees	0	2	0
Parameter-Efficient Fine-Tuning Enhances Adaptation of Single Cell Large Language Model for Cell Type Identification · DOI	Agrees	1	0	0

Comparisons in Wu et al. 2025 (4)

Protocol	Dataset	Metric	scGPT	Geneformer	Higher
Cross-atlas annotation transfer, HLCA to Tabula Sapiens, 14 shared cell types (Wu et al. 2025 Table S3)	Tabula Sapiens (CELLxGENE), 120 leaf cell types after processing	Macro F1 (test cells of the 14 shared cell types; averaged over 'all classes' (Methods 'Macro-F1'), not stated as the 14 shared types or every source-atlas type), fraction	0.209	0.239	Geneformer
Cross-atlas annotation transfer, HLCA to Tabula Sapiens, 14 shared cell types (Wu et al. 2025 Table S3)	Tabula Sapiens (CELLxGENE), 120 leaf cell types after processing	Top-1 accuracy (Accuracy@1 over cells of the 14 shared cell types), fraction	0.614	0.772	Geneformer
Cross-atlas annotation transfer, Tabula Sapiens to HLCA, 14 shared cell types (Wu et al. 2025 Table S2)	HLCA core (CELLxGENE), 37 leaf cell types after processing	Macro F1 (test cells of the 14 shared cell types; averaged over 'all classes' (Methods 'Macro-F1'), not stated as the 14 shared types or every source-atlas type), fraction	0.081	0.087	Geneformer
Cross-atlas annotation transfer, Tabula Sapiens to HLCA, 14 shared cell types (Wu et al. 2025 Table S2)	HLCA core (CELLxGENE), 37 leaf cell types after processing	Top-1 accuracy (Accuracy@1 over cells of the 14 shared cell types), fraction	0.674	0.665	scGPT

Comparisons in Kedzierska et al. 2025 (97)

Protocol	Dataset	Metric	scGPT	Geneformer	Higher
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (overall average ranking across datasets, based on mean biological-conservation and batch-mixing scores (1 = best); full embedding dimensions), unitless	4	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (rank of the median ARI score across datasets among the eight representations (1 = best); full embedding dimensions), unitless	4	7	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (rank of the median ASW score across datasets among the eight representations (1 = best); full embedding dimensions), unitless	3	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (rank of the median Average BIO score across datasets among the eight representations (1 = best); full embedding dimensions), unitless	4	7	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (rank of the median Average batch score across datasets among the eight representations (1 = best); full embedding dimensions), unitless	1	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (rank of the median Batch integration score across datasets among the eight representations (1 = best); full embedding dimensions), unitless	2	7	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (rank of the median NMI score across datasets among the eight representations (1 = best); full embedding dimensions), unitless	4	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (rank of the median PCR score across datasets among the eight representations (1 = best); full embedding dimensions), unitless	1	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.4	0.19	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions), unitless	0.91	0.38	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions), unitless	0.53	0.35	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.66	0.42	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions), unitless	2	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions), unitless	0.89	0.76	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions), unitless	0.52	0.45	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Immune (330k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions), unitless	0.95	0	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.34	0.09	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions), unitless	0.94	0.3	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions), unitless	0.44	0.24	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.5	0.26	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions), unitless	4	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions), unitless	0.96	0.6	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions), unitless	0.49	0.37	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (16k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions), unitless	0.93	0	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (8k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.14	0.05	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (8k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions), unitless	0.94	0.36	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (8k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions), unitless	0.35	0.22	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (8k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.4	0.16	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (8k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions), unitless	0.96	0.73	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (8k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions), unitless	0.5	0.46	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Pancreas (8k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions), unitless	0.92	0	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.89	0.6	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions), unitless	0.84	0.48	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions), unitless	0.78	0.59	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.85	0.65	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions), unitless	3	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions), unitless	0.97	0.97	Tie
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions), unitless	0.61	0.51	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (12k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions), unitless	0.71	0	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (95k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.46	0.4	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (95k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions), unitless	0.55	0.52	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (95k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.65	0.65	Tie
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (95k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions), unitless	3	6	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	PBMC (95k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions), unitless	0.56	0.51	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.31	0.14	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions), unitless	0.9	0.45	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions), unitless	0.45	0.3	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions), unitless	0.56	0.35	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions), unitless	2	8	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions), unitless	0.92	0.9	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions), unitless	0.47	0.41	scGPT
Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions), unitless	0.91	0	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Evaluated datasets, aggregated (Kedzierska et al. 2025)	Rank (overall average ranking across datasets, based on mean biological-conservation and batch-mixing scores (1 = best); UMAP projection of the embedding), unitless	4	8	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.35	0.12	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding), unitless	0.76	0.24	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding), unitless	0.48	0.29	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.6	0.34	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding), unitless	4	8	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding), unitless	0.6	0.49	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding), unitless	0.5	0.4	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Immune (330k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding), unitless	0.96	0	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.26	0.07	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding), unitless	0.9	0.18	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding), unitless	0.4	0.22	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.46	0.24	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding), unitless	3	8	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding), unitless	0.82	0.37	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding), unitless	0.48	0.35	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (16k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding), unitless	0.98	0	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (8k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.13	0.02	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (8k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding), unitless	0.88	0.26	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (8k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding), unitless	0.29	0.2	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (8k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.37	0.16	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (8k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding), unitless	0.8	0.53	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (8k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding), unitless	0.43	0.42	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Pancreas (8k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding), unitless	0.95	0	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.57	0.37	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding), unitless	0.93	0.71	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding), unitless	0.68	0.51	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.75	0.58	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding), unitless	3	7	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding), unitless	0.96	0.89	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding), unitless	0.74	0.6	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (12k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding), unitless	0.9	0.53	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (95k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.43	0.3	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (95k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding), unitless	0.54	0.45	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (95k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.59	0.48	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (95k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding), unitless	3	6	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	PBMC (95k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding), unitless	0.61	0.58	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.33	0.11	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding), unitless	0.8	0.35	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding), unitless	0.41	0.22	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding), unitless	0.54	0.3	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding), unitless	3	8	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding), unitless	0.69	0.71	Geneformer
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding), unitless	0.37	0.26	scGPT
Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Tabula Sapiens (483k) (Kedzierska et al. 2025)	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding), unitless	0.91	0	scGPT

Comparisons in Deep-learning-based gene perturbation… (2)

Protocol	Dataset	Metric	scGPT	Geneformer	Higher
Norman double perturbations: expression prediction for held-out double perturbations (Ahlmann-Eltze et al. 2025 Fig. 1a and Extended Data Fig. 2a)	Norman et al. 2019 Perturb-seq, K562 CRISPRa single and double perturbations (scFoundation reprocessing, Ahlmann-Eltze et al. 2025)	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits), unit-unreported	4.62	3.48	Geneformer
Norman double perturbations: expression prediction for held-out double perturbations (Ahlmann-Eltze et al. 2025 Fig. 1a and Extended Data Fig. 2a)	Norman et al. 2019 Perturb-seq, K562 CRISPRa single and double perturbations (scFoundation reprocessing, Ahlmann-Eltze et al. 2025)	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits), unitless	0.527	0.773	Geneformer

Comparisons in Parameter-Efficient Fine-Tuning… (1)

Protocol	Dataset	Metric	scGPT	Geneformer	Higher
Cell-type identification	M.S. single-cell dataset	F1, unitless	0.734	0.388	scGPT

What disagrees

Other studies: rank scGPT above Geneformer in 95 of 98 comparisons
This study: ranks Geneformer above scGPT in 3 of 4 comparisons

Compared: Geneformer, scGPT
Study that disagrees: Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13)
Other studies: 8 studies
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx), Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101), Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines, Source Data Extended Data Fig. 2, Ahlmann-Eltze et al. 2025, Source Data Extended Data Fig. 8, Ahlmann-Eltze et al. 2025, Source Data Fig. 1, Ahlmann-Eltze et al. 2025, Source Data Fig. 2, Ahlmann-Eltze et al. 2025, Parameter-Efficient Fine-Tuning Enhances Adaptation of Single Cell Large Language Model for Cell Type Identification
Rule: Ordering reversal; different tasks across studies
Model identity: Matched across studies partly through links not yet reviewed

The compared numbers

Each value links to its result record and source. Compare values within one protocol; the rule compares rankings, not raw values across protocols.

In the study that disagrees (8)
Study	Tool	Protocol or benchmark	Metric	Value
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	Geneformer zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, HLCA to Tabula Sapiens, 14 shared cell types (Wu et al. 2025 Table S3)	Top-1 accuracy (Accuracy@1 over cells of the 14 shared cell types)	0.772
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	Geneformer zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, HLCA to Tabula Sapiens, 14 shared cell types (Wu et al. 2025 Table S3)	Macro F1 (test cells of the 14 shared cell types; averaged over 'all classes' (Methods 'Macro-F1'), not stated as the 14 shared types or every source-atlas type)	0.239
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	scGPT zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, HLCA to Tabula Sapiens, 14 shared cell types (Wu et al. 2025 Table S3)	Top-1 accuracy (Accuracy@1 over cells of the 14 shared cell types)	0.614
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	scGPT zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, HLCA to Tabula Sapiens, 14 shared cell types (Wu et al. 2025 Table S3)	Macro F1 (test cells of the 14 shared cell types; averaged over 'all classes' (Methods 'Macro-F1'), not stated as the 14 shared types or every source-atlas type)	0.209
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	Geneformer zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, Tabula Sapiens to HLCA, 14 shared cell types (Wu et al. 2025 Table S2)	Top-1 accuracy (Accuracy@1 over cells of the 14 shared cell types)	0.665
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	Geneformer zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, Tabula Sapiens to HLCA, 14 shared cell types (Wu et al. 2025 Table S2)	Macro F1 (test cells of the 14 shared cell types; averaged over 'all classes' (Methods 'Macro-F1'), not stated as the 14 shared types or every source-atlas type)	0.087
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	scGPT zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, Tabula Sapiens to HLCA, 14 shared cell types (Wu et al. 2025 Table S2)	Top-1 accuracy (Accuracy@1 over cells of the 14 shared cell types)	0.674
Wu et al. 2025, Additional file 3 (Supplementary Tables S1-S13); Biology-driven insights into the power of single-cell foundation models	scGPT zero-shot embeddings with OnClass (Wu et al. 2025)	Cross-atlas annotation transfer, Tabula Sapiens to HLCA, 14 shared cell types (Wu et al. 2025 Table S2)	Macro F1 (test cells of the 14 shared cell types; averaged over 'all classes' (Methods 'Macro-F1'), not stated as the 14 shared types or every source-atlas type)	0.081

In the other studies (207 results)

In the other studies
Study	Tool	Protocol or benchmark	Metric	Value
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (overall average ranking across datasets, based on mean biological-conservation and batch-mixing scores (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median ARI score across datasets among the eight representations (1 = best); full embedding dimensions)	7
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median ASW score across datasets among the eight representations (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median Average batch score across datasets among the eight representations (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median Average BIO score across datasets among the eight representations (1 = best); full embedding dimensions)	7
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median Batch integration score across datasets among the eight representations (1 = best); full embedding dimensions)	7
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median NMI score across datasets among the eight representations (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median PCR score across datasets among the eight representations (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (overall average ranking across datasets, based on mean biological-conservation and batch-mixing scores (1 = best); full embedding dimensions)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median Average BIO score across datasets among the eight representations (1 = best); full embedding dimensions)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median NMI score across datasets among the eight representations (1 = best); full embedding dimensions)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median ARI score across datasets among the eight representations (1 = best); full embedding dimensions)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median ASW score across datasets among the eight representations (1 = best); full embedding dimensions)	3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median Average batch score across datasets among the eight representations (1 = best); full embedding dimensions)	1
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median Batch integration score across datasets among the eight representations (1 = best); full embedding dimensions)	2
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank of the median PCR score across datasets among the eight representations (1 = best); full embedding dimensions)	1
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.19
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.45
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.38
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.35
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.76
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.42
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.52
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.53
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.89
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.66
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.52
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.89
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.91
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0.95
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	2
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.09
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.37
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.24
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.6
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.26
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.49
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.34
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.49
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.44
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.5
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.49
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.94
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.96
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0.93
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.05
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.46
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.36
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.22
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.73
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.16
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.14
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.5
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.35
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.94
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.96
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0.92
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.6
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.51
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.48
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.59
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.65
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.89
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.61
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.78
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.85
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.84
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0.71
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.51
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.52
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	6
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.56
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.46
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.55
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.14
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.41
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.45
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.9
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.35
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.31
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.47
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; full embedding dimensions)	0.45
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; full embedding dimensions)	0.92
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; full embedding dimensions)	0.56
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); full embedding dimensions)	2
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib asw label (cell-type ASW; whole dataset; full embedding dimensions)	0.47
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; full embedding dimensions)	0.9
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring, full embedding dimensions	Scib pcr batch (principal component regression on batch; whole dataset; full embedding dimensions)	0.91
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (overall average ranking across datasets, based on mean biological-conservation and batch-mixing scores (1 = best); UMAP projection of the embedding)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (overall average ranking across datasets, based on mean biological-conservation and batch-mixing scores (1 = best); UMAP projection of the embedding)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.12
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.24
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.29
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.49
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.34
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.35
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.5
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.48
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.6
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.35
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT kidney (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.6
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.76
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0.96
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.07
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.35
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.18
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.22
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.37
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.24
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.26
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.48
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.4
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.46
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.9
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.82
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0.98
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.02
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.42
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.26
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.2
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.53
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.16
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.13
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.29
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.37
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT kidney (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.43
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.43
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.88
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0.95
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.37
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.6
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.71
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.51
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.89
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.58
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0.53
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	7
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.57
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.68
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.75
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.74
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.93
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.96
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0.9
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.58
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.45
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.48
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	6
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.43
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.61
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.54
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.59
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.11
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.26
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.35
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.22
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.71
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	Geneformer (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw label (cell-type ASW; whole dataset; UMAP projection of the embedding)	0.37
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Avg bio (AvgBIO (mean of cell-type ASW, NMI and ARI; NMI and ARI on Louvain clusters at the best-NMI resolution); whole dataset; UMAP projection of the embedding)	0.41
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT blood (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Normalized mutual information (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.54
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Adjusted rand index (Louvain clusters at best-NMI resolution versus cell-type labels; whole dataset; UMAP projection of the embedding)	0.33
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT human (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Rank (rank among the eight representations by mean biological-conservation and batch-mixing scores on this dataset (1 = best); UMAP projection of the embedding)	3
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT kidney (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.69
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Average batch score (average batch score (mean of batch integration score and PCR batch score); whole dataset; UMAP projection of the embedding)	0.8
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib asw batch (batch integration score (batch ASW per cell type); whole dataset; UMAP projection of the embedding)	0.69
Kedzierska et al. 2025, Additional file 1 (Supplementary Tables S1-S7, xlsx); Zero-shot evaluation reveals limitations of single-cell foundation models (Kedzierska et al. 2025, Genome Biology 26:101)	scGPT, randomly initialised (Kedzierska et al. 2025)	Kedzierska et al. 2025 cell embedding scoring after UMAP projection	Scib pcr batch (principal component regression on batch; whole dataset; UMAP projection of the embedding)	0.91
Parameter-Efficient Fine-Tuning Enhances Adaptation of Single Cell Large Language Model for Cell Type Identification	scGPT	Cell-type identification	F1	0.734
Parameter-Efficient Fine-Tuning Enhances Adaptation of Single Cell Large Language Model for Cell Type Identification	Geneformer	Cell-type identification	F1	0.388
Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines; Source Data Extended Data Fig. 2, Ahlmann-Eltze et al. 2025; Source Data Fig. 1, Ahlmann-Eltze et al. 2025	Geneformer (Ahlmann-Eltze et al. 2025)	Norman double perturbations: expression prediction for held-out double perturbations (Ahlmann-Eltze et al. 2025 Fig. 1a and Extended Data Fig. 2a)	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	0.773
Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines; Source Data Extended Data Fig. 2, Ahlmann-Eltze et al. 2025; Source Data Fig. 1, Ahlmann-Eltze et al. 2025	scGPT (Ahlmann-Eltze et al. 2025)	Norman double perturbations: expression prediction for held-out double perturbations (Ahlmann-Eltze et al. 2025 Fig. 1a and Extended Data Fig. 2a)	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	0.527
Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines; Source Data Fig. 1, Ahlmann-Eltze et al. 2025	Geneformer (Ahlmann-Eltze et al. 2025)	Norman double perturbations: expression prediction for held-out double perturbations (Ahlmann-Eltze et al. 2025 Fig. 1a and Extended Data Fig. 2a)	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	3.48
Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines; Source Data Fig. 1, Ahlmann-Eltze et al. 2025	scGPT (Ahlmann-Eltze et al. 2025)	Norman double perturbations: expression prediction for held-out double perturbations (Ahlmann-Eltze et al. 2025 Fig. 1a and Extended Data Fig. 2a)	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	4.62

Follow-up

Current state: Not yet reviewed.

No one has reviewed this finding yet.

How to help resolve this

If you know why these studies disagree, for example a different population, metric or preprocessing step, or you have a corrected value, tell us through the contribution form. A reviewer checks every submission against its source before anything changes, and the step is added to the follow-up above.

Finding discrepancy-1980d1e1d257348a · All contested results · How findings are made and followed up