© 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. 40218
    Prev
    Next
    INDEX
    Explanations

    Locations US, favourite blog, suicide bombers

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

    The neuron consistently lights up on unusually long tokens (e.g. long words or multi‐letter identifiers), effectively detecting tokens that exceed a typical length threshold.

    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
     what
    -2.52
     how
    -1.94
     as
    -1.91
     they
    -1.90
     always
    -1.76
     using
    -1.76
     that
    -1.74
     possibly
    -1.72
     which
    -1.71
     not
    -1.69
    POSITIVE LOGITS
    llllll
    1.65
    ครื่อง
    1.62
    titudes
    1.62
     énon
    1.59
     offre
    1.59
     faciliter
    1.57
    และ
    1.54
     déplo
    1.54
     สำหรับ
    1.53
     kitab
    1.53
    Activations Density 0.265%

    No Known Activations