© 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. 22-GEMMASCOPE-RES-131K
    7. 14777
    Prev
    Next
    INDEX
    Explanations

    programming literals and types

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

    The neuron fires strongly on tokens involved in Python string/text‐type handling (e.g. text_type, ensure_str, decode, ascii, isinstance checks), so it’s detecting code that deals with converting or ensuring proper text/string encoding.

    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
     ję
    -1.02
     continue
    -0.96
     户型
    -0.95
     słoń
    -0.94
     trasero
    -0.93
    ğim
    -0.93
     zwierzę
    -0.93
     τὸν
    -0.92
     skór
    -0.91
     żel
    -0.90
    POSITIVE LOGITS
     Dasar
    1.22
     tekno
    1.16
     seksi
    1.04
     kasar
    1.04
     Presiden
    1.02
    ങ്ങ
    1.00
    //'
    0.99
     kanak
    0.98
     pandemi
    0.98
     kalah
    0.98
    Activations Density 0.010%

    No Known Activations