© 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. 25512
    Prev
    Next
    INDEX
    Explanations

    project introduction

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

    The neuron is triggered by long runs of leading whitespace (i.e. heavy indentation) in code or comment blocks.

    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
    राब
    -1.06
     individuel
    -0.94
    ———-
    -0.91
     diinginkan
    -0.91
     corporativa
    -0.90
     peint
    -0.89
    cée
    -0.89
    lllllll
    -0.88
    INIS
    -0.87
     BIL
    -0.86
    POSITIVE LOGITS
    ることができます
    0.94
    atihan
    0.84
    losse
    0.82
    aggi
    0.81
    errHandler
    0.79
    0.79
     whom
    0.79
    break
    0.77
     grants
    0.75
     waaronder
    0.75
    Activations Density 0.002%

    No Known Activations