© 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. 78325
    Prev
    Next
    INDEX
    Explanations

    form submission

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

    The neuron strongly activates on programming‐language tokens—especially HTML attributes, event‐binding directives, identifiers like “form” or component names, and other code symbols—i.e. code snippet elements.

    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
    rati
    -0.86
     Raceway
    -0.85
     proyección
    -0.84
     deschis
    -0.81
     utilidad
    -0.81
     choisir
    -0.80
     saída
    -0.79
     MatDialog
    -0.79
    ⠦
    -0.78
    paghetti
    -0.77
    POSITIVE LOGITS
     onSubmit
    2.42
    onSubmit
    2.28
     handleSubmit
    2.02
    handleSubmit
    1.84
     submit
    1.69
     submission
    1.54
    submit
    1.39
    Submit
    1.30
     Submission
    1.28
     Submit
    1.26
    Activations Density 0.004%

    No Known Activations