© 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. 22-GEMMASCOPE-RES-131K
    7. 128086
    Prev
    Next
    INDEX
    Explanations

    The main thing this neuron does is pick out high-intensity evaluative adjectives and adverbs (especially positive, marketing-style descriptors).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    positive descriptive qualities

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     rique
    -1.73
     maravillas
    -1.69
     Erziehung
    -1.68
     cenar
    -1.67
     concursos
    -1.63
     السياس
    -1.48
     exportaciones
    -1.47
     preocupaciones
    -1.46
     reír
    -1.45
    Neden
    -1.44
    POSITIVE LOGITS
     philippe
    1.41
     monter
    1.38
     THURSDAY
    1.35
     is
    1.34
     kiel
    1.31
     设计
    1.30
     archetype
    1.30
     fritt
    1.29
     TUESDAY
    1.29
     pengar
    1.28
    Activations Density 0.053%

    No Known Activations