© 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. 10-GEMMASCOPE-RES-131K
    7. 85690
    Prev
    Next
    INDEX
    Explanations

    JSON and code structures

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

    This neuron activates strongly on leading-indent whitespace (the sequences of spaces/tabs used for indentation).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     should
    -2.33
     to
    -2.30
     will
    -2.17
     like
    -1.98
     with
    -1.96
     or
    -1.92
     at
    -1.90
     time
    -1.88
     for
    -1.84
     Though
    -1.84
    POSITIVE LOGITS
     parfü
    2.14
     stratég
    2.06
    侂
    2.03
    絎
    1.98
    鵙
    1.95
     peinado
    1.95
     continúas
    1.95
     exámenes
    1.92
    쑹
    1.90
    皤
    1.88
    Activations Density 0.113%

    No Known Activations