Home Â» Database

AVID-2026-R0135

Description

The model Kimi-K2-Instruct-0905 from Moonshot AI was evaluated by the Garak LLM Vulnerability scanner using the probe divergence.RepeatedToken.

Details

The probe divergence.RepeatedToken tests if a model leaks data by repeating a string multiple times. This probe is part of the divergence module which detects divergence in model output.

The LLM Kimi-K2-Instruct-0905 was evaluated on this probe.

Metrics

Number of failures were counted for each detector in the probe.

DetectorPassedTotalScore
divergence.RepeatedToken54256🟠 21.1

References

Affected or Relevant Artifacts

  • Developer: Moonshot AI
  • Deployer: Together AI
  • Artifact Details:
TypeName
ModelKimi-K2-Instruct-0905

Impact

AVID Taxonomy Categorization

  • Risk domains: Performance
  • SEP subcategories: P0403: Socioeconomic safety
  • Lifecycle stages: L05: Evaluation

Other information

  • Report Type: Measurement
  • Credits:
  • Date Reported: 2026-02-20
  • Version: 0.3.1
  • AVID Entry