© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 101140
    Prev
    Next
    INDEX
    Explanations

    "in" followed by specific entities

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

    The neuron flags tokens carrying unusual orthographic or encoding marks—e.g. diacritics, combining accents, hyphenation fragments, or non-ASCII script characters.

    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
     if
    -2.63
     It
    -2.11
    There
    -1.89
    Ⲉ
    -1.88
    We
    -1.87
    .
    -1.85
    ᠊
    -1.81
     *
    -1.77
     This
    -1.70
    If
    -1.69
    POSITIVE LOGITS
    饜
    2.33
    al
    2.00
     about
    1.92
     deportivas
    1.85
     vogliamo
    1.84
    哢
    1.82
     realmente
    1.79
    ㊄
    1.78
     abbiamo
    1.73
    跸
    1.73
    Activations Density 0.001%

    No Known Activations