© 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. 66138
    Prev
    Next
    INDEX
    Explanations

    derogatory terms

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

    The neuron activates on highly negative, insulting words and slurs—i.e. derogatory epithets and abusive language.

    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
     about
    -1.54
     erfordert
    -1.52
     –
    -1.41
     its
    -1.41
    v
    -1.36
     also
    -1.35
     where
    -1.29
    contentInset
    -1.28
     her
    -1.21
    価格は
    -1.20
    POSITIVE LOGITS
    ัตร
    1.32
    CELONA
    1.28
    prä
    1.24
     Перейти
    1.20
    anschluss
    1.19
     jednom
    1.17
    wedding
    1.16
    drage
    1.16
    sprecher
    1.15
    providedIn
    1.14
    Activations Density 0.062%

    No Known Activations