© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 99678
    Prev
    Next
    INDEX
    Explanations

    public class `class`

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

    This neuron fires on code-structure and documentation markers—primarily XML doc tags (e.g. “<”, “>”, “summary” delimiters) and braces (“{”/“}”)—i.e. syntactic delimiters in source files.

    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
     Nasıl
    -0.92
    zeption
    -0.88
     chatting
    -0.84
    妹
    -0.82
     pulver
    -0.82
    江
    -0.81
     miRNAs
    -0.81
     לח
    -0.80
     Hora
    -0.80
    Meg
    -0.80
    POSITIVE LOGITS
     delitos
    0.93
    ⩽
    0.82
     beserta
    0.82
    +}$
    0.81
     ounces
    0.81
     koos
    0.81
    是非常
    0.79
    ﺰ
    0.78
    ֩
    0.77
    ὴν
    0.76
    Activations Density 0.012%

    No Known Activations