© 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. 10-GEMMASCOPE-RES-131K
    7. 14423
    Prev
    Next
    INDEX
    Explanations

    common starting fragments of sentences

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

    The neuron is picking up section or paragraph separators (i.e. blank‐line or divider tokens that mark breaks between chunks of text).

    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
     cráneo
    -2.02
     it
    -1.98
    你先
    -1.96
     siento
    -1.95
    简称
    -1.94
     basado
    -1.93
    靣
    -1.83
    <bos>
    -1.81
     tuve
    -1.79
    你别
    -1.77
    POSITIVE LOGITS
     cucchia
    1.97
     are
    1.84
     RISERV
    1.83
    🩹
    1.74
    ？！”
    1.71
     such
    1.70
     GONNA
    1.68
    ของ
    1.65
    maktadır
    1.63
     necessaria
    1.63
    Activations Density 0.017%

    No Known Activations