© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 113372
    Prev
    Next
    INDEX
    Explanations

    safety and security

    np_acts-logits-general · gemini-2.5-flash-lite

    words related to security or safety across multiple languages (particularly "sicurezza" in Italian, "sécurité" in French, "seguridad" in Spanish, "bezpečnost" in Czech, and "secure" in English).

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    This neuron responds to words (in various languages) containing the root for “security” or “safety” (e.g. secur-, sicher-, sikker-, sicur-).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    )
    -2.66
    y
    -2.17
    \
    -2.11
    ;
    -1.99
     When
    -1.99
    .”
    -1.95
    </strong>
    -1.93
    啬
    -1.93
     причиной
    -1.92
     encapsulated
    -1.91
    POSITIVE LOGITS
    EnableWeb
    2.73
    跻
    2.23
    鈊
    2.22
    綻
    2.19
    before
    2.19
    猹
    2.17
    鱘
    2.11
    酝
    2.08
    pinephrine
    2.06
    逦
    2.06
    Activations Density 0.006%

    No Known Activations