© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 12269
    Prev
    Next
    INDEX
    Explanations

    Vue component or academic abstract

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

    The neuron fires strongly on code‐style punctuation and syntax markers (dots, parentheses, braces, etc.), effectively detecting programming language 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
     Dima
    -0.79
    ctrons
    -0.77
     werk
    -0.73
    expre
    -0.73
     Stry
    -0.72
     عمده
    -0.72
     angles
    -0.71
     expr
    -0.71
     Angles
    -0.71
    onTap
    -0.69
    POSITIVE LOGITS
    refs
    1.01
    Skład
    0.79
    transition
    0.79
    HelloWorld
    0.79
     Transition
    0.77
    lement
    0.76
     grandchild
    0.75
    Inne
    0.74
     Vee
    0.73
    Pretty
    0.73
    Activations Density 0.026%

    No Known Activations