© 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. 34-GEMMASCOPE-RES-131K
    7. 78443
    Prev
    Next
    INDEX
    Explanations

    copyright notices and company names

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

    The neuron strongly activates on the common boilerplate words (“the,” “with,” “and,” “or,” etc.) found in source‐code copyright and license header blocks.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     larger
    -0.84
     Disadvantages
    -0.82
     setImage
    -0.81
    anything
    -0.80
     too
    -0.79
     setType
    -0.79
    ユニクロ
    -0.79
     Vå
    -0.79
     disadvantage
    -0.77
     seda
    -0.77
    POSITIVE LOGITS
     vertrouwen
    0.94
    ֺ
    0.90
     kraju
    0.89
     viagem
    0.86
     venganza
    0.85
     aguard
    0.84
     serisi
    0.84
    ˗
    0.83
    lected
    0.82
    jük
    0.82
    Activations Density 0.406%

    No Known Activations