© 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. 1862
    Prev
    Next
    INDEX
    Explanations

    Taia or Ta ta

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

    The neuron strongly activates on tokens that begin with the two‐letter sequence “Ta,” i.e. proper‐name or technical prefixes starting with “Ta.”

    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
    圌
    -3.48
    ’，
    -3.09
     ,‬
    -2.92
     internacionales
    -2.81
    䊐
    -2.75
     médicas
    -2.70
    踽
    -2.64
     tüm
    -2.58
    大量的
    -2.56
     sostener
    -2.53
    POSITIVE LOGITS
    In
    2.69
    *
    2.67
     It
    2.66
     Its
    2.55
     the
    2.50
    For
    2.41
    <
    2.41
    To
    2.41
    A
    2.39
    There
    2.38
    Activations Density 0.004%

    No Known Activations