© 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. 34-GEMMASCOPE-RES-131K
    7. 3161
    Prev
    Next
    INDEX
    Explanations

    special characters and identifiers

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

    The neuron strongly fires on repeated occurrences of the “+=” token (i.e. lines made up of long runs of the += operator).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    erdings
    -1.02
     while
    -0.88
     or
    -0.86
     Türk
    -0.85
     time
    -0.85
     several
    -0.84
     mín
    -0.84
     kval
    -0.83
     sampah
    -0.82
     negatif
    -0.82
    POSITIVE LOGITS
    before
    0.83
     имеется
    0.81
     magna
    0.80
    other
    0.79
    jména
    0.78
     recie
    0.77
    keep
    0.76
     foll
    0.76
     übrigens
    0.75
     înainte
    0.75
    Activations Density 0.004%

    No Known Activations