© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 78621
    Prev
    Next
    INDEX
    Explanations

    killing, death, destruction

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

    The neuron fires strongly on words naming or referring to lethal violence (e.g. kill, murder, slay).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     eso
    -0.81
     وكان
    -0.80
    newPassword
    -0.75
    pausal
    -0.74
     Penrose
    -0.74
     estis
    -0.74
    –
    -0.73
     energia
    -0.73
    skys
    -0.72
     einer
    -0.71
    POSITIVE LOGITS
     στοι
    0.92
     потенциаль
    0.88
     killed
    0.88
     proportionality
    0.88
    をご覧ください
    0.85
     erziel
    0.81
    chap
    0.79
     supplémentaire
    0.78
    pacho
    0.75
     element
    0.75
    Activations Density 0.038%

    No Known Activations