rewirebio.iobenchmarks
Configuration

SCimilarity (HyCell report)

Produces a frozen cell embedding with SCimilarity as the HyCell report runs it.

14 evaluations · 14 results

Overview

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

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

Model Capabilities & Hardware Requirements

Deployment facts recorded for SCimilarity (HyCell report). Fields the record does not carry are marked not recorded.

ParametersNot recorded
Minimum VRAMNot recorded
Inference LatencyNot recorded
Context WindowNot recorded
AccessNot recorded
LicenseNot recorded

Evaluated Checkpoint: Released weights at default inference settings; no revision printed

Evaluations and results

14 evaluations · 14 results. Different protocols are not a single leaderboard.

Sorted by Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases) (higher is better). The best value in each column is highlighted. Decimals are rounded for display; each value links to the printed value and its source.

Tested configurationProtocol and datasetWeighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)Macro F1 (Macro-F1 of a 15-nearest-neighbour classifier trained on the other batches, over the cell types of the held-out batch, averaged over held-out batches; per atlas or averaged over the seven atlases)Macro F1 (Macro-F1 of a logistic-regression classifier trained on the other batches, over the cell types of the held-out batch, averaged over held-out batches; per atlas or averaged over the seven atlases)Scib batch correction score (Mean of five scIB batch-correction metrics (batch silhouette, iLISI, kBET, graph connectivity and principal-component regression), per atlas or averaged over the seven atlases)Scib bio conservation score (Mean of seven scIB biological-conservation metrics (isolated-label score, NMI and ARI of k-means and of Leiden clusters against the cell types, cell-type silhouette and cLISI), per atlas or averaged over the seven atlases)Spearman (Spearman correlation between diffusion pseudotime and differentiation stage (naive B, germinal-centre B, plasmablast, plasma cell) over the 3,751 cells of these states, rooted at the medoid of the 50 cells with the highest naive-B marker score (TCL1A, IGHD, FCER2))Spearman (The same Spearman correlation between pseudotime and stage computed per donor and averaged over the 12 donors)
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), PBMC from two 10x Genomics runs (HyCell report)0.848 (best in column)Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), MultiVINE brain vasculature, healthy donors (HyCell report)0.781Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), scIB human immune atlas, integer-count part (HyCell report)0.729Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), Seven held-out atlases of the HyCell report0.726Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), scIB human lung atlas, integer-count part (HyCell report)0.684Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), Asian Immune Diversity Atlas, sample (HyCell report)0.682Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), Cross-tissue immune atlas, stratified sample (HyCell report)0.68Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationComposite score S (HyCell report), scIB human pancreas atlas, integer-count part (HyCell report)0.677Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationAnnotation transfer with 15-nearest neighbours (HyCell report), Seven held-out atlases of the HyCell reportNot reported0.791Not reportedNot reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationAnnotation transfer with logistic regression (HyCell report), Seven held-out atlases of the HyCell reportNot reportedNot reported0.839Not reportedNot reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationscIB batch correction (HyCell report), Seven held-out atlases of the HyCell reportNot reportedNot reportedNot reported0.542Not reportedNot reportedNot reported
SCimilarity (HyCell report)EvaluationscIB biological conservation (HyCell report), Seven held-out atlases of the HyCell reportNot reportedNot reportedNot reportedNot reported0.699Not reportedNot reported
SCimilarity (HyCell report)EvaluationB-cell to plasma-cell pseudotime ordering, donor mean (HyCell report), Cross-tissue immune atlas, stratified sample (HyCell report)Not reportedNot reportedNot reportedNot reportedNot reportedNot reported0.868
SCimilarity (HyCell report)EvaluationB-cell to plasma-cell pseudotime ordering, all cells (HyCell report), Cross-tissue immune atlas, stratified sample (HyCell report)Not reportedNot reportedNot reportedNot reportedNot reported0.908Not reported
All 14 result rows with coverage, uncertainty and sources
Filter evaluations

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configurationProtocol and datasetFindingEvidence and details
Configuration: SCimilarity (HyCell report)Protocol: Annotation transfer with 15-nearest neighbours (HyCell report)
Dataset: Seven held-out atlases of the HyCell report
0.791 Macro F1 (Macro-F1 of a 15-nearest-neighbour classifier trained on the other batches, over the cell types of the held-out batch, averaged over held-out batches; per atlas or averaged over the seven atlases)
fraction · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on annotation-transfer-knn, seven-held-out-atlases (HyCell report)

hycell-20261106-protocol-annotation-transfer-knn

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 3, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'kNN' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Annotation transfer with logistic regression (HyCell report)
Dataset: Seven held-out atlases of the HyCell report
0.839 Macro F1 (Macro-F1 of a logistic-regression classifier trained on the other batches, over the cell types of the held-out batch, averaged over held-out batches; per atlas or averaged over the seven atlases)
fraction · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on annotation-transfer-lr, seven-held-out-atlases (HyCell report)

hycell-20261106-protocol-annotation-transfer-lr

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 3, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'LR' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: scIB batch correction (HyCell report)
Dataset: Seven held-out atlases of the HyCell report
0.542 Scib batch correction score (Mean of five scIB batch-correction metrics (batch silhouette, iLISI, kBET, graph connectivity and principal-component regression), per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on batch-correction, seven-held-out-atlases (HyCell report)

hycell-20261106-protocol-batch-correction

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 3, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Batch' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: scIB biological conservation (HyCell report)
Dataset: Seven held-out atlases of the HyCell report
0.699 Scib bio conservation score (Mean of seven scIB biological-conservation metrics (isolated-label score, NMI and ARI of k-means and of Leiden clusters against the cell types, cell-type silhouette and cLISI), per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on bio-conservation, seven-held-out-atlases (HyCell report)

hycell-20261106-protocol-bio-conservation

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 3, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Bio' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: Asian Immune Diversity Atlas, sample (HyCell report)
0.682 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, aida (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 4, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'AIDA' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: MultiVINE brain vasculature, healthy donors (HyCell report)
0.781 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, brain (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 4, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Brain' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: Cross-tissue immune atlas, stratified sample (HyCell report)
0.68 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, cross-tissue-immune (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 4, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Cross-tissue' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: scIB human immune atlas, integer-count part (HyCell report)
0.729 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, immune (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 4, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Immune' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: scIB human lung atlas, integer-count part (HyCell report)
0.684 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, lung (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 4, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Lung' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: scIB human pancreas atlas, integer-count part (HyCell report)
0.677 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, pancreas (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 4, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Pancreas' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: PBMC from two 10x Genomics runs (HyCell report)
0.848 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, pbmc (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 4, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'PBMC' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: Composite score S (HyCell report)
Dataset: Seven held-out atlases of the HyCell report
0.726 Weighted composite score (S = 0.3 Bio + 0.2 Batch + 0.25 LR + 0.25 kNN, the authors' weighting, which gives the scIB total (0.6 Bio + 0.4 Batch) and annotation transfer equal weight; per atlas or averaged over the seven atlases)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on composite-score, seven-held-out-atlases (HyCell report)

hycell-20261106-protocol-composite-score

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Table 3, row 'SCimilarityl' (printed 'SCimilarity' with superscript l, trained with cell-type labels), column 'Score S' (page 5 of 9)
Configuration: SCimilarity (HyCell report)Protocol: B-cell to plasma-cell pseudotime ordering, donor mean (HyCell report)
Dataset: Cross-tissue immune atlas, stratified sample (HyCell report)
0.868 Spearman (The same Spearman correlation between pseudotime and stage computed per donor and averaged over the 12 donors)
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on pseudotime-stage-spearman-donor-mean, cross-tissue-immune (HyCell report)

hycell-20261106-protocol-pseudotime-stage-spearman-donor-mean

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Section 4.2 (page 7 of 9), averaged over donors, SCimilarity
Configuration: SCimilarity (HyCell report)Protocol: B-cell to plasma-cell pseudotime ordering, all cells (HyCell report)
Dataset: Cross-tissue immune atlas, stratified sample (HyCell report)
0.908 Spearman (Spearman correlation between diffusion pseudotime and differentiation stage (naive B, germinal-centre B, plasmablast, plasma cell) over the 3,751 cells of these states, rooted at the medoid of the 50 cells with the highest naive-B marker score (TCL1A, IGHD, FCER2))
unitless · higher

Uncertainty: Not reported by the source: No uncertainty is printed anywhere in the report.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

SCimilarity on pseudotime-stage-spearman, cross-tissue-immune (HyCell report)

hycell-20261106-protocol-pseudotime-stage-spearman

Aggregation: Not reported

HyCell: a Compact Embedding Model for Single-Cell Transcriptomes (Hyra Team, 2026-11-01) · Section 4.2 (page 7 of 9), all cells, SCimilarity

Source checking is not independent reproduction. Release 2026-11-21-c8bea25e5abb.

Use this configuration

How it works, versions and access

Additional methodological details, versions and access terms have not yet been extracted into a structured profile for this configuration. Consult the linked sources and evaluation records below.

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

0 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-11-21-c8bea25e5abb
Property and statementOriginal source and locationReview and provenance

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

Sources and history

Release 2026-11-21-c8bea25e5abb · Record review: source checked

1 source record and release historyDownload this release (gzip)
Technical metadata and record identifiers

Technical metadata for this record is tracked under stable ID hycell-20261106-config-scimilarity.

Source locator: Section 3.1 'Models'; Table 3

Related records

How to cite

SCimilarity (HyCell report). Rewire biological model benchmark database, release 2026-11-21-c8bea25e5abb. https://benchmarks.rewirebio.io/database/configuration/hycell-20261106-config-scimilarity/

Release 2026-11-21-c8bea25e5abb. Rankings and values change between releases, so keep the release id in the citation. Cite the original publication as well for any value you reuse: each result links its source.

The compilation and annotations made by Rewire in this release are offered under CC BY 4.0; values and text taken from cited sources stay under the terms of those sources and are not covered by this licence. Data licence

Suggest a correction