© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 68406
    Prev
    Next
    INDEX
    Explanations

    code validation logic

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

    The neuron fires mainly on code tokens that name or invoke functions/methods (e.g. identifiers before “(” or “.”), effectively picking out function/method definitions and calls.

    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
    äls
    -1.07
     LAC
    -0.93
    GLFW
    -0.92
    acruz
    -0.90
    mMap
    -0.89
     reinigen
    -0.88
     arcade
    -0.85
    cocos
    -0.82
    痉
    -0.81
     动物
    -0.81
    POSITIVE LOGITS
     new
    0.89
     choosing
    0.83
    ={<
    0.82
     zamiast
    0.82
    examples
    0.81
     menyebabkan
    0.80
    codable
    0.78
     centavos
    0.77
    bouncycastle
    0.77
     talented
    0.76
    Activations Density 0.043%

    No Known Activations