© 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. 22-GEMMASCOPE-RES-131K
    7. 73071
    Prev
    Next
    INDEX
    Explanations

    import statements and code

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

    The neuron strongly activates on structural code and markup tokens (e.g. closing brackets/tags, punctuation-heavy identifiers, and attribute delimiters) rather than natural-language words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     discuter
    -1.03
     famí
    -0.98
    viève
    -0.98
     escaping
    -0.97
     rapidement
    -0.94
     berbagai
    -0.93
    strando
    -0.93
    urare
    -0.91
    FieldOffset
    -0.90
     vocale
    -0.89
    POSITIVE LOGITS
     with
    0.94
     medium
    0.91
     moderately
    0.90
     цьому
    0.79
     женский
    0.79
     Saludos
    0.78
     Moderately
    0.78
     smaller
    0.77
    呈現
    0.77
     règne
    0.77
    Activations Density 0.014%

    No Known Activations