© 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. 72549
    Prev
    Next
    INDEX
    Explanations

    tap, drag, hold, click

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

    The neuron fires on words naming user input gestures or events (e.g. “double‐click,” “tap,” “swipe,” “hold,” “rotate”).

    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
     alcune
    -1.16
     molte
    -1.14
     certaines
    -1.10
     niektórych
    -1.10
     alcuni
    -1.10
    those
    -1.09
     mogą
    -1.09
    seems
    -1.05
    嬸
    -1.05
    aéroport
    -1.04
    POSITIVE LOGITS
     or
    1.43
     at
    1.30
     to
    1.23
    ;->
    1.16
     before
    1.09
     یا
    1.07
     on
    1.05
    入院
    1.00
     longer
    0.99
     through
    0.95
    Activations Density 0.013%

    No Known Activations