memory benchmark
CI Memories Violation Rate
CI Memories measures privacy failures in memory-enabled agents using contextual-integrity scenarios. This metric is the violation rate; lower is better.
Updated Aug 11, 2026
Models1
Model coverage1
MetricScore
EvidenceB
CI Memories Violation Rate Ranking
Higher score ranks better on this benchmark.
| 01 | ME | 26.4% | 100.0% | 1 | C |
CI Memories Violation Rate Highlights
The leading models and scores on this benchmark.
What is CI Memories Violation Rate?
What CI Memories Violation Rate measures and how its scores work.
CI Memories measures privacy failures in memory-enabled agents using contextual-integrity scenarios. This metric is the violation rate; lower is better.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- CI Memories Violation Rate
- Modality
- text
- Primary category
- memory
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- ci-memories-violation|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about CI Memories Violation Rate.
Which model scores highest on CI Memories Violation Rate?
Muse Glimmer-30B is currently ranked first with 26.4%.
What does CI Memories Violation Rate measure?
CI Memories measures privacy failures in memory-enabled agents using contextual-integrity scenarios. This metric is the violation rate; lower is better.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.