google-deepmind / shieldgemma-2-4b-it

ShieldGemma 2 is a model trained on Gemma 3's 4B IT checkpoint for image safety classification across key categories that takes in images and outputs safety labels per policy.

  • Public
  • 210 runs
  • L40S
  • GitHub
  • Weights
  • Paper
  • License
  • Prediction

    google-deepmind/shieldgemma-2-4b-it:d6db0e17215ac8e04db2a8edd78195ae811a6d98ef8152f5a9f56787ec7f7c4a
    ID
    8n5regs5adrme0cnn97trmy784
    Status
    Succeeded
    Source
    Web
    Hardware
    L40S
    Total duration
    Created

    Input

    image
    image
    policy_type
    dangerous_content

    Output

    { "probabilities": { "no": 0.9999998807907104, "yes": 1.1067029248579274e-7 } }
    Generated in
  • Prediction

    google-deepmind/shieldgemma-2-4b-it:d6db0e17215ac8e04db2a8edd78195ae811a6d98ef8152f5a9f56787ec7f7c4a
    ID
    dz48zvj7f1rmc0cnn9b8r2yje0
    Status
    Succeeded
    Source
    Web
    Hardware
    L40S
    Total duration
    Created

    Input

    image
    image
    policy_type
    violence_gore

    Output

    { "probabilities": { "no": 1, "yes": 8.361025760628138e-11 } }
    Generated in
  • Prediction

    google-deepmind/shieldgemma-2-4b-it:d6db0e17215ac8e04db2a8edd78195ae811a6d98ef8152f5a9f56787ec7f7c4a
    ID
    c5chx4k68hrma0cnn9b99dya50
    Status
    Succeeded
    Source
    Web
    Hardware
    L40S
    Total duration
    Created

    Input

    image
    image
    policy_type
    sexually_explicit

    Output

    { "probabilities": { "no": 0.999996542930603, "yes": 0.000003494503289402928 } }
    Generated in

Want to make some of these yourself?

Run this model