PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
covariate transfer on Srivatsan20, RMSE of the mean. Scored with RMSE of the mean on Srivatsan20. Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds.
Overview
covariate transfer on Srivatsan20, RMSE of the mean. Scored with RMSE of the mean on Srivatsan20. Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds.
Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.
Evaluation design
Benchmarks bring together tasks and protocols. A task describes the biological question; a protocol defines a particular test.
Benchmarks
These source-backed links do not make different protocols or scores interchangeable.
Recorded evaluations
Each evaluation records what was tested and under which conditions.
- Biolord* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- CPA* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- CPA* (noAdv) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- CPA* (scGPT) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- Decoder (Cov) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- Decoder on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- LA on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- LA (scGPT) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- Linear on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- SAMS-VAE* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
Run instructions
No runnable recipe has been reviewed for this task. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.
A task describes a biological question. Choose a linked protocol to obtain concrete split and scoring instructions.
Published comparisons
Explore the results reported under one evaluation protocol. Each figure keeps its source, dataset and metric together; it is not a ranking across studies. The pooled view gathers every source table that reports the same metric and names what it does not hold constant.
PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
rmse_mean (error) · Lower values are better for this metric.
Every method PerturBench reports on covariate transfer on Srivatsan20, RMSE of the mean, scored with RMSE of the mean on Srivatsan20.
- CPA* · Configuration · Author-reported evaluation0.021 ± 4 × 10 − 4
- CPA* (noAdv) · Configuration · Author-reported evaluation0.020 ± 8 × 10 − 4
- CPA* (scGPT) · Configuration · Author-reported evaluation0.021 ± 3 × 10 − 4
- SAMS-VAE* · Configuration · Author-reported evaluation0.060 ± 1 × 10 − 3
- Biolord* · Configuration · Author-reported evaluation0.086 ± 4 × 10 − 2
- LA · Configuration · Author-reported evaluation0.018 ± 6 × 10 − 5
- LA (scGPT) · Configuration · Author-reported evaluation0.017 ± 1 × 10 − 4
- Decoder · Configuration · Author-reported evaluation0.018 ± 1 × 10 − 4
- Decoder (Cov) · Configuration · Author-reported evaluation0.023 ± 3 × 10 − 5
- Linear · Configuration · Author-reported evaluation0.030 ± 5 × 10 − 4
Source order is preserved. Plotted marks show point estimates; uncertainty, where reported, is retained in the printed values and table. Differences do not establish statistical significance.
PerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, column(RMSE, mean)Values, uncertainty and evidence
| Tested entity | Printed value | Uncertainty | Evidence |
|---|---|---|---|
| CPA* · Configuration | 0.021 ± 4 × 10 − 4 error | type: standard_deviation; value: 0.0004 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(CPA ∗), column(RMSE, mean) |
| CPA* (noAdv) · Configuration | 0.020 ± 8 × 10 − 4 error | type: standard_deviation; value: 0.0008 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(CPA ∗ (noAdv)), column(RMSE, mean) |
| CPA* (scGPT) · Configuration | 0.021 ± 3 × 10 − 4 error | type: standard_deviation; value: 0.0003 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(CPA ∗ (scGPT)), column(RMSE, mean) |
| SAMS-VAE* · Configuration | 0.060 ± 1 × 10 − 3 error | type: standard_deviation; value: 0.001 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(SAMS-VAE ∗), column(RMSE, mean) |
| Biolord* · Configuration | 0.086 ± 4 × 10 − 2 error | type: standard_deviation; value: 0.04 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Biolord ∗), column(RMSE, mean) |
| LA · Configuration | 0.018 ± 6 × 10 − 5 error | type: standard_deviation; value: 0.00006 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(LA), column(RMSE, mean) |
| LA (scGPT) · Configuration | 0.017 ± 1 × 10 − 4 error | type: standard_deviation; value: 0.0001 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(LA (scGPT)), column(RMSE, mean) |
| Decoder · Configuration | 0.018 ± 1 × 10 − 4 error | type: standard_deviation; value: 0.0001 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Decoder), column(RMSE, mean) |
| Decoder (Cov) · Configuration | 0.023 ± 3 × 10 − 5 error | type: standard_deviation; value: 0.00003 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Decoder (Cov)), column(RMSE, mean) |
| Linear · Configuration | 0.030 ± 5 × 10 − 4 error | type: standard_deviation; value: 0.0005 | Author-reported evaluation · source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Linear), column(RMSE, mean) |
Scope and limitations
- Author-reported numbers, source checked but not independently reproduced.
- RMSE and both rank metrics are better when lower. The rank metrics measure how often another perturbation's prediction is closer than the right one's.
- The two experiments use different datasets and splits, so their figures are not comparable to each other.
Source transcription and grouping reviewed by automated source review on 2026-09-18. These experiments were not independently reproduced by rewire.
Tested entities and results
Release 2026-09-17-134cd1815de8 · 10 evaluations · 10 metric rows. Different protocols are not a single leaderboard. Where several source tables report the same metric, the published comparisons above offer a pooled view that names what it does not hold constant.
| Metric and finding | Coverage and uncertainty | Evidence |
|---|---|---|
| Biolord* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: Biolord*Task: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.086 ± 4 × 10 − 2 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.04 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Biolord ∗), column(RMSE, mean) Source checking is not independent reproduction. |
| CPA* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: CPA*Task: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.021 ± 4 × 10 − 4 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.0004 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(CPA ∗), column(RMSE, mean) Source checking is not independent reproduction. |
| CPA* (noAdv) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: CPA* (noAdv)Task: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.020 ± 8 × 10 − 4 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.0008 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(CPA ∗ (noAdv)), column(RMSE, mean) Source checking is not independent reproduction. |
| CPA* (scGPT) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: CPA* (scGPT)Task: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.021 ± 3 × 10 − 4 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.0003 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(CPA ∗ (scGPT)), column(RMSE, mean) Source checking is not independent reproduction. |
| Decoder (Cov) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: Decoder (Cov)Task: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.023 ± 3 × 10 − 5 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.00003 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Decoder (Cov)), column(RMSE, mean) Source checking is not independent reproduction. |
| Decoder on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: DecoderTask: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.018 ± 1 × 10 − 4 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.0001 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Decoder), column(RMSE, mean) Source checking is not independent reproduction. |
| LA on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: LATask: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.018 ± 6 × 10 − 5 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.00006 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(LA), column(RMSE, mean) Source checking is not independent reproduction. |
| LA (scGPT) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: LA (scGPT)Task: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.017 ± 1 × 10 − 4 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.0001 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(LA (scGPT)), column(RMSE, mean) Source checking is not independent reproduction. |
| Linear on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: LinearTask: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.030 ± 5 × 10 − 4 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.0005 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(Linear), column(RMSE, mean) Source checking is not independent reproduction. |
| SAMS-VAE* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean Configuration: SAMS-VAE*Task: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the meanDataset subset: Srivatsan20 (PerturBench split) Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds. Author-reported evaluation · Evaluation metadata: source checked | ||
| 0.060 ± 1 × 10 − 3 rmse_mean Unit: error · Direction: lower | Uncertainty: type: standard deviation; value: 0.001 Scored: Not reported · Eligible: Not reported | source checkedPerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Table 2, row(SAMS-VAE ∗), column(RMSE, mean) Source checking is not independent reproduction. |
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
1 evidence row matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Relationship: part of discovery-benchmark-perturbench Individual claims | PerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis Table 2, column(RMSE, mean) Version: Primary full-text snapshot retrieved 2026-09-17; exact bytes pinned by SHA-256 | source checked automated source review · 2026-09-18 Audit detailsPrimary-source transcription with no human sign-off and no independent reproduction. Field: Claim: perturbench-association-ct-rmse Source artifact SHA-256: Hash scope: Exact retrieved primary paper artifact bytes. |
Sources and history
Release 2026-09-17-134cd1815de8 · Record review: source checked
1 source records and release history
- PerturBench: Benchmarking Machine Learning Models for Cellular Perturbation Analysis · Original source · Primary full-text snapshot retrieved 2026-09-17; exact bytes pinned by SHA-256
Technical metadata and extraction receipts
Stable ID: perturbench-task-ct-rmse
- areas
- cells-tissues
- tasks
- covariate transfer on Srivatsan20, RMSE of the mean
- metric
- RMSE of the mean
- metric direction
- lower
- dataset
- Srivatsan20
- protocol
- Train on some cell types and predict drug effects in a held-out cell type, reported as the mean and one standard deviation over seeds.
- source locator
- Table 2, column(RMSE, mean)
- comparison panels
- id: perturbench-panel-ct-rmse; title: PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean; protocol id: perturbench-task-ct-rmse; dataset id: perturbench-dataset-srivatsan20; metric: rmse_mean; unit: error; direction: lower; result ids: perturbench-result-cpa-ct-rmse-rmse-mean; perturbench-result-cpa-noadv-ct-rmse-rmse-mean; perturbench-result-cpa-scgpt-ct-rmse-rmse-mean; perturbench-result-sams-vae-ct-rmse-rmse-mean; perturbench-result-biolord-ct-rmse-rmse-mean; perturbench-result-la-ct-rmse-rmse-mean; perturbench-result-la-scgpt-ct-rmse-rmse-mean; perturbench-result-decoder-ct-rmse-rmse-mean; perturbench-result-decoder-cov-ct-rmse-rmse-mean; perturbench-result-linear-ct-rmse-rmse-mean; source ids: evidence-expansion-perturbench-5c480456; source locator: Table 2, column(RMSE, mean); context: Every method PerturBench reports on covariate transfer on Srivatsan20, RMSE of the mean, scored with RMSE of the mean on Srivatsan20.; caveats: Author-reported numbers, source checked but not independently reproduced.; RMSE and both rank metrics are better when lower. The rank metrics measure how often another perturbation's prediction is closer than the right one's.; The two experiments use different datasets and splits, so their figures are not comparable to each other.; review: method: automated_source_review; date: 2026-09-18
Related records
- part of: PerturBench
- subject: PerturBench CT-RMSE: part of discovery-benchmark-perturbench
- benchmark: Biolord* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: CPA* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: CPA* (noAdv) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: CPA* (scGPT) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: Decoder (Cov) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: Decoder on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: LA on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: LA (scGPT) on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: Linear on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean
- benchmark: SAMS-VAE* on PerturBench CT-RMSE: covariate transfer on Srivatsan20, RMSE of the mean