Muse Spark 1.1 Evaluation Report
Meta's evaluation report for Muse Spark 1.1, dated July 9, 2026.
- Issued by
- Meta
- Authors credited
- No authorship claim recorded yet. The source record lists Meta.
- Dates
- Published 2026-07-09 issue evidence
The report prints 'Date: July 9, 2026', day precision, without saying what kind of date it is. No revision date is printed in the copy read.
- Exact source
- Muse Spark 1.1 Evaluation Report Meta · published 2026-07-09 · Retrieved for the candidate pack on 2026-11-05 (SHA-256 cd25634e0c8a3960d47dadfe6f49eefd3c12f6f62ad0dfbdf1cb4c96b5b5ade9) directly from the publisher’s host. Retrieved again on 2026-11-05, after the adjudication, with the same SHA-256; the cross-family check of 2026-11-05 retrieved the same SHA-256.
Appears in
No collection holds a record about this document yet.
Versions
meta-muse-spark-1-1-evaluation-report: 1 recorded version, oldest first, published by Meta. A version is shown as replacing another only where a record says so. Date order alone never implies it.
- 2026-07-09 · publishedAs dated July 9, 2026 (this version)
This version
Provisions in this version · 15 quoted · 1 not found
A limited selection of quoted provisions: what this edition says about thresholds, required actions, who decides and exceptions, and other statements of the text. It is not a full summary and says nothing about implementation. A “not found” record states which part of the edition was inspected. It is not a claim that the text was deleted.
- Model-card questionCard date“Date: July 9, 2026”Opening report summary, PDF p. 2 (unnumbered)Inspect claim & evidence ↗
- Model-card questionRelease date“Muse Spark 1.1 is also released to external developers through an application programming interface (API)”§1 Introduction, PDF p. 6, printed p. 5Inspect claim & evidence ↗
- Model-card questionFramework cited“Meta’s Advanced AI Scaling Framework”§1 Introduction, footnote 2, PDF p. 6, printed p. 5Inspect claim & evidence ↗
- Model-card questionLink to framework“https://ai.meta.com/static-resource/Meta_Advanced-AI-Scaling-Framework-v2”§1, footnote 2, PDF p. 6, printed p. 5Inspect claim & evidence ↗
- Model-card questionFramework determination“residual risk across the Chemical & Biological, Cybersecurity, and Loss of Control domains is “moderate or lower” following mitigations.”§1.1.1 Risk Assessment Summary, PDF p. 8, printed p. 7Inspect claim & evidence ↗
- Model-card questionSafeguards tied to determination“Our determination that the deployment of Muse Spark 1.1, if left unmitigated, could substantially contribute to catastrophic outcomes related to Chemical & Biological risks requires us to design, implement, and validate mitigations”§2.1 Chemical & Biological Risks, Mitigations and Safeguards, PDF p. 12, printed p. 11Inspect claim & evidence ↗
- Model-card questionChemical and biological“We therefore retain the “high risk” determination for the unmitigated model under Outcomes CB-1 and CB-2.”§2.1 Chemical & Biological Risks, Summary of Results, PDF p. 12, printed p. 11Inspect claim & evidence ↗
- Model-card questionCyber“Our evaluations cannot rule out a “high risk” designation for the unmitigated model in the Cybersecurity domain under our Advanced AI Scaling Framework.”§2.2 Cybersecurity, Capability Assessment and Summary of Results, PDF p. 19, printed p. 18Inspect claim & evidence ↗
- Model-card questionAutonomy and AI R&D“Muse Spark 1.1 still lacks sufficient autonomy to sustain automated AI R&D”§2.3.1.1 Acceleration of AI Development, Results, PDF p. 41, printed p. 40Inspect claim & evidence ↗
- Model-card questionPersuasion“We assess the Cyber 3 (scaled fraud, extortion, and scams) outcome through a multi-turn social-engineering capability evaluation”§2.2.2 Social Engineering, PDF p. 27, printed p. 26Inspect claim & evidence ↗
- Model-card questionAlignment and deception“The presence of the reasoning therefore has no measurable effect on whether the model actually complies.”§4.1.5 Scheming, Alignment Faking results, PDF p. 65, printed p. 64Inspect claim & evidence ↗
- Model-card questionRed-teaming“GraySwan additionally ran prompt-injection attacks using their private Agent Red Teaming (ART) benchmark”§3.2.2.3 Third-Party Red Teaming: GraySwan ART, PDF p. 56, printed p. 55Inspect claim & evidence ↗
- Model-card questionExternal testers“Scale AI runs the private split of FORTRESS and we include the results in Figure 26.”§3.1.3.1 FORTRESS, PDF p. 50, printed p. 49Inspect claim & evidence ↗
- Model-card questionRelease decision“we therefore release Muse Spark 1.1.”Opening summary, PDF p. 2 (unnumbered)Inspect claim & evidence ↗
- Model-card questionStated limits“Because examples were adversarially curated or generated, this evaluation does not measure how frequently sycophancy occurs in average user conversations”§4.1.4 Sycophancy, PDF p. 63, printed p. 62Inspect claim & evidence ↗
- Not found in the inspected textGovernment instituteInspected: Muse Spark 1.1 Evaluation Report: §§1–1.4, §§2–4 third-party evaluation references and bibliography; whole-report searchInspect claim & evidence ↗
Earliest version recorded here; no replacement is recorded.
Issue and authorship records
Documentary relations
Each relation is listed separately. Citing, crediting and reporting the use of a method are different acts, and none of them shows agreement or influence by itself.
Cites · 1
Explicit references in this document. Citation is neither agreement nor influence.
Not recorded for this document: cited by · explicitly credits · reports using a method from · method reported as used by · applies parts of · parts applied by · discusses · reports evaluation of. A missing record does not show that the relation does not exist.