© 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. 10-GEMMASCOPE-RES-131K
    7. 58654
    Prev
    Next
    INDEX
    Explanations

    curse words and profanity

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

    The neuron fires on strong profane and slur terms—most intensely on the N-word and other harsh expletives.

    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
     bonitos
    -2.17
    </strong>
    -1.88
    余额
    -1.87
    臯
    -1.87
    announced
    -1.86
     ساله
    -1.81
     medlems
    -1.80
     nasze
    -1.78
    采用了
    -1.77
    书籍
    -1.77
    POSITIVE LOGITS
     Thus
    2.20
     Therefore
    2.16
     termed
    2.03
     Importantly
    1.95
    皷
    1.89
     Incidentally
    1.84
     Indeed
    1.84
     研究
    1.79
     Moreover
    1.76
     denoting
    1.73
    Activations Density 0.007%

    No Known Activations