© 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. 128687
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on code tokens naming or referencing dialog/modal UI components (e.g. “modal”, “Modal”, “Dialog”, “AlertDialog”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    dialogs and alerts

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     glimmer
    -0.93
    цыі
    -0.90
    াপ
    -0.89
     бъ
    -0.86
    Fence
    -0.85
    柔らか
    -0.85
     Czę
    -0.84
    ſt
    -0.84
     &
    -0.83
    Etat
    -0.82
    POSITIVE LOGITS
     explained
    1.20
     explain
    1.20
     asked
    1.13
     forklar
    1.13
    ventajas
    1.05
    ご了承
    1.04
    neiden
    1.02
     wanted
    0.99
     explains
    0.97
     wollte
    0.96
    Activations Density 0.012%

    No Known Activations