© 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. 10-GEMMASCOPE-RES-131K
    7. 15347
    Prev
    Next
    INDEX
    Explanations

    programming code constructs

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

    The neuron fires on the “.” token used for object or module member access (e.g. object.property or module.function).

    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
     is
    -2.17
     of
    -2.08
     which
    -2.06
     it
    -1.97
     just
    -1.96
     In
    -1.93
     only
    -1.91
     this
    -1.89
     also
    -1.77
     the
    -1.73
    POSITIVE LOGITS
     некоторы
    1.95
    实话
    1.84
    begins
    1.80
    varies
    1.73
    appears
    1.73
    馊
    1.71
     работают
    1.71
     certains
    1.69
    twelve
    1.66
    ༑
    1.65
    Activations Density 0.040%

    No Known Activations