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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 87163
    Prev
    Next
    INDEX
    Explanations

    devil and satan mentions

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

    The neuron fires strongly on words that denote mighty or supernatural actions (e.g. expel, cast out, defeat, divide).

    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
    atherapy
    -0.77
    英雄
    -0.77
     Tourist
    -0.76
    унд
    -0.74
    Migrate
    -0.74
    Ignored
    -0.73
     chrys
    -0.71
     Favre
    -0.69
     Cine
    -0.69
     Bress
    -0.69
    POSITIVE LOGITS
     devil
    4.91
     Devil
    4.31
    Devil
    3.88
     Satan
    3.88
    devil
    3.48
    Satan
    3.25
     devils
    3.08
     satan
    2.97
     Devils
    2.77
    魔鬼
    2.44
    Activations Density 0.016%

    No Known Activations