Model type
Cell representation and perturbation-transition model family
Model
STATE
State separates cellular representation learning from prediction of responses to perturbation.
Sources
ArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and LicensesResults are available for configurations using this model. Their fitted heads, extra inputs and evaluation settings are kept separate below.
How it worksSTATE workflow
Conceptual summary of the documented data flow; optional inputs and configured downstream stages must be reported for a reproducible evaluation.
Sources
ArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and LicensesOverview
Inputs
AnnData expression measurements, gene/cell-type labels and an explicit dataset/split configuration.
Outputs
Cell embeddings or predicted perturbed-expression matrices, depending on the selected component.
Access
Official project documentation and implementation: https://github.com/ArcInstitute/state
Sources
ArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenseslimited source coverage · Automated source review, 2026-09-16. All specifications and missing details
Related configurations, pipelines and services
These configurations, services and pipelines use this model within their own configurations. Their results, where available, are not assigned to the underlying model.
- Maksim Bobrin: STATE protein transfer v1 [submission 8VtfKKCtALf6cDkh4fqB] Configuration · 1 result
- LME: Model-v11.1-ArcState-AdaptiveGating [submission M9DkSYMrvCMqlp2lbfRX] Configuration · 1 result
- Khanh Nguyen: Arc-STATE-baseline-v1 [submission jACV1avJMLtZvNBU61Uj] Configuration · 1 result
- Lijiayu Deng: State-GenePT-v001-panel-counts [submission vUSQaMGSF6bXuCMEYu18] Configuration · 1 result
Use this model
How it works, versions and access
How it works
How it works
State separates cellular representation learning from prediction of responses to perturbation. State Embedding and State Transition are separate components; the documented transition workflow trains on specified expression features and perturbation metadata. The documented inputs are annData expression measurements, gene/cell-type labels and an explicit dataset/split configuration. The output consists of cell embeddings or predicted perturbed-expression matrices, depending on the selected component.
Sources
ArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and LicensesVersions and reproducibility
State Embedding (SE) and State Transition (ST), with separately trained checkpoints. The applicable input limits require configuration-specific checking.
Sources
ArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and LicensesStrengths, limitations and unresolved questions
Strengths and limitations
Strengths and considerations
- The command-line workflow makes training, inference and evaluation configurations explicit.
Sources
ArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses
Limitations and conditions
- State Embedding and State Transition are not interchangeable model identities. Data omitted from zero-shot/few-shot split declarations default to training in the documented configuration format.
Sources
ArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses
Profile review details
Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied.
Stable record: discovery-model-stateSpecifications
Inputs, training, access and other details
Explanatory profile: limited source coverage · Automated source review, 2026-09-16. Review applies to the cited claims; unresolved fields are listed below. Numerical results retain their own review status.
| Property | Description and evidence |
|---|---|
| Model type | Cell representation and perturbation-transition model familySourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Architecture | State Embedding and State Transition are separate components; the documented transition workflow trains on specified expression features and perturbation metadata.SourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Inputs | AnnData expression measurements, gene/cell-type labels and an explicit dataset/split configuration.SourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Outputs | Cell embeddings or predicted perturbed-expression matrices, depending on the selected component.SourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Parameters | The repository contains separate State Embedding and State Transition configurations; the selected complete pipeline is required before a parameter total can be assigned. · Not reported in inspected sourcesSourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Known versions | State Embedding (SE) and State Transition (ST), with separately trained checkpoints.SourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Training data | Examples include Replogle–Nadig genetic perturbations and Tahoe-100M; the actual trained model depends on its configuration.SourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Training cutoff | The reviewed README and training configuration do not provide one latest-data date shared by all State Embedding and task-fitted State Transition releases. · Not reported in inspected sourcesSourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Context limits | The transition workflow operates on the chosen expression features and perturbation metadata; the inspected family documentation does not define one universal gene/cell token budget. · Not reported in inspected sourcesSourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Weights licence | Arc Research Institute State Model Non-Commercial License, with Acceptable Use Policy; separate from code CC-BY-NC-SA-4.0.SourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Access | Official project documentation and implementation: https://github.com/ArcInstitute/stateSourcesArcInstitute/state: README.md · README.md: Getting started, State Transition Model, ST TOML configuration and Licenses |
| Code licence | CC-BY-NC-SA-4.0SourcesArcInstitute/state: LICENSE · LICENSE: licence text |
Applicable tests and references
Applicability is distinct from a completed evaluation.
- Virtual Cell Challenge 2026 · Proposed association
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.
20 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Diagram caption Conceptual summary of the documented data flow; optional inputs and configured downstream stages must be reported for a reproducible evaluation. Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
Diagram steps
| ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Diagram title STATE workflow Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Model type Cell representation and perturbation-transition model family Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Architecture State Embedding and State Transition are separate components; the documented transition workflow trains on specified expression features and perturbation metadata. Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Access Official project documentation and implementation: https://github.com/ArcInstitute/state Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Code licence CC-BY-NC-SA-4.0 Individual claims | ArcInstitute/state: LICENSE LICENSE: licence text Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Inputs AnnData expression measurements, gene/cell-type labels and an explicit dataset/split configuration. Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Outputs Cell embeddings or predicted perturbed-expression matrices, depending on the selected component. Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | source checked automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
| Parameters The repository contains separate State Embedding and State Transition configurations; the selected complete pipeline is required before a parameter total can be assigned. Individual claims | ArcInstitute/state: README.md README.md: Getting started, State Transition Model, ST TOML configuration and Licenses Version: 9bbfe78a434a55205e4de834e1ea99f85f7a3add | unreported automated source review · 2026-09-16 Audit detailsInspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied. Field: Source artifact SHA-256: Hash scope: SHA-256 of retrieved original artifact bytes Format: original_artifact |
Sources and history
View linked audit checks and correction history
Release 2026-09-29-06401fd5b220 · Record review: discovered
3 source records and release history
- ArcInstitute/state official source · Original source · 9bbfe78a434a55205e4de834e1ea99f85f7a3add
- ArcInstitute/state: README.md · Original source · 9bbfe78a434a55205e4de834e1ea99f85f7a3add
- ArcInstitute/state: LICENSE · Original source · 9bbfe78a434a55205e4de834e1ea99f85f7a3add
Technical metadata and extraction receipts
Stable ID: discovery-model-state
- areas
- single-cell
- access
- official_source_linked
- benchmark applicability
- candidate; not evidence of a reported evaluation
- candidate benchmark ids
- discovery-benchmark-virtual-cell-challenge-2026
- entity level
- family
- reported name
- STATE
- version
- Not reported
- historical missing metadata
- checkpoint: unextracted; code licence: unextracted; parameters: unextracted; training cutoff: unextracted; training data: unextracted; version: unextracted; weights licence: unextracted
- metadata review scope
- historical_missing_metadata preserves the original discovery state. Current descriptive evidence and missingness are recorded in profile.facts; numerical-result review is separate.
- entity classification
- review date: 2026-09-17; rationale: The cited profile describes a named learned biological predictor or representation model/family. Preserve this identity separately from task-specific fitting, individual checkpoints, pipelines and hosted access.; source ids: evidence-official-1df5b1865861177c0c75; source locator: README.md: Getting started, State Transition Model, ST TOML configuration and Licenses; ambiguities: None recorded
Related records
- applicable to: Virtual Cell Challenge 2026
- uses model: Maksim Bobrin: STATE protein transfer v1 [submission 8VtfKKCtALf6cDkh4fqB]
- uses model: LME: Model-v11.1-ArcState-AdaptiveGating [submission M9DkSYMrvCMqlp2lbfRX]
- uses model: Khanh Nguyen: Arc-STATE-baseline-v1 [submission jACV1avJMLtZvNBU61Uj]
- uses model: Lijiayu Deng: State-GenePT-v001-panel-counts [submission vUSQaMGSF6bXuCMEYu18]