© 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. 57920
    Prev
    Next
    INDEX
    Explanations

    code delimiters

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

    This neuron primarily fires on HTML tag elements—especially the “<” bracket and tag names like form, div, input—indicating it’s detecting HTML markup.

    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
    ihnachts
    -0.93
    結
    -0.87
    idavit
    -0.85
    aurait
    -0.81
     MÁS
    -0.81
     pomoci
    -0.79
     ужа
    -0.78
    athered
    -0.78
    ihnachten
    -0.77
    mogelijk
    -0.77
    POSITIVE LOGITS
    ="{{
    0.74
    abri
    0.72
    annon
    0.71
    richter
    0.69
    utar
    0.69
     drown
    0.69
    圳
    0.68
    ínio
    0.68
    四年
    0.67
     like
    0.66
    Activations Density 0.006%

    No Known Activations