© 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. 79116
    Prev
    Next
    INDEX
    Explanations

    cookie, script, expires

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

    The neuron fires whenever it spots programming‐style syntax—things like function/var keywords, parentheses, braces, semicolons, and other code tokens.

    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
    dafri
    -0.81
     行
    -0.79
    寄
    -0.78
     við
    -0.77
     Agustus
    -0.77
    pick
    -0.76
     عطر
    -0.76
    bastian
    -0.76
    stof
    -0.75
    ￡
    -0.74
    POSITIVE LOGITS
    続く
    0.90
     Seulement
    0.89
    besos
    0.81
     podrob
    0.80
    使った
    0.80
     Storing
    0.77
     réputation
    0.76
    紆
    0.76
    Savannah
    0.75
    SEGU
    0.75
    Activations Density 0.010%

    No Known Activations