© 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. 26710
    Prev
    Next
    INDEX
    Explanations

    constructor

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

    The neuron fires on programming identifiers—particularly variable or field names (often containing underscores) in code.

    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
     hunne
    -1.55
    だけではなく
    -1.42
     feroit
    -1.41
     enlever
    -1.41
    ごち
    -1.41
    regarding
    -1.34
     Provides
    -1.33
     présents
    -1.32
     erhi
    -1.31
     améli
    -1.30
    POSITIVE LOGITS
     in
    1.32
    Even
    1.30
     Even
    1.28
     when
    1.23
    When
    1.21
     most
    1.20
     even
    1.20
     will
    1.20
    Some
    1.20
    ！！
    1.20
    Activations Density 0.009%

    No Known Activations