© 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. 10-GEMMASCOPE-RES-131K
    7. 48133
    Prev
    Next
    INDEX
    Explanations

    evil

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

    The neuron strongly detects occurrences of the word “evil” (and its close morphological variants).

    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
     totalling
    -1.62
    ੴ
    -1.61
     grouped
    -1.59
    豋
    -1.58
    以上です
    -1.57
    过来的
    -1.57
    ・・・」
    -1.57
    
    -1.57
    I
    -1.56
    . 
    -1.55
    POSITIVE LOGITS
     Mientras
    1.96
    7
    1.84
    也
    1.82
     Aunque
    1.74
    }
    1.73
     The
    1.70
     โดย
    1.68
     neuen
    1.66
     rzeczy
    1.66
     také
    1.66
    Activations Density 0.032%

    No Known Activations