© 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. 31380
    Prev
    Next
    INDEX
    Explanations

    code keywords or separators

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

    The neuron activates on boilerplate “code‐generated” header comments (e.g. “// Code generated … DO NOT EDIT”).

    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
    degrees
    -1.06
    -0.84
    ස
    -0.82
     sopa
    -0.81
     auxiliar
    -0.81
    fills
    -0.81
     え
    -0.80
    İM
    -0.79
    
    -0.79
    diad
    -0.79
    POSITIVE LOGITS
    了自己
    0.94
    這些
    0.85
    textrm
    0.81
     tiến
    0.79
     erste
    0.79
     dessa
    0.78
    гда
    0.77
     New
    0.77
     حيث
    0.77
     القرار
    0.76
    Activations Density 0.016%

    No Known Activations