© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-4B-IT
    5. 22-GEMMASCOPE-2-RES-16K
    7. 456
    Prev
    Next
    INDEX
    Explanations

    the model's refusal to generate harmful, unethical, or inappropriate content.

    oai_token-act-pair · gemini-2.5-flash
    Triggered by @ananthanv94
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-4b-it/resid_post/layer_22_width_16k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     gives
    0.58
     provides
    0.58
     denotes
    0.57
     merupakan
    0.55
     adalah
    0.54
    。
    0.53
     Provides
    0.53
     constitutes
    0.52
     generates
    0.52
     requires
    0.52
    POSITIVE LOGITS
    However
    1.05
    But
    0.96
    Therefore
    0.95
    Consequently
    0.84
    但是
    0.84
    BUT
    0.83
    therefore
    0.81
    Furthermore
    0.80
    Moreover
    0.79
     However
    0.78
    Activations Density 1.199%

    No Known Activations