© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 130613
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on descriptive adjectives—especially color words (e.g. “red,” “white”), nationalities or breeds (“German,” “shepherd”), and other attributive adjectives (“longitudinal,” “reciprocal”).

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

    colors, nationalities, or scientific terms

    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
    deserved
    -0.93
     fritt
    -0.93
    chọn
    -0.88
     Ninh
    -0.85
    restaurants
    -0.84
     puits
    -0.84
     commerces
    -0.84
     significantly
    -0.83
     snabb
    -0.83
     Dames
    -0.82
    POSITIVE LOGITS
    aucune
    0.93
     when
    0.93
     through
    0.91
     or
    0.88
     may
    0.86
    したり
    0.85
    What
    0.85
     billeder
    0.84
     Perhaps
    0.84
    墩
    0.81
    Activations Density 0.108%

    No Known Activations