© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 15985
    Prev
    Next
    INDEX
    Explanations

    networking

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

    This neuron consistently fires on the very first meaningful word of each new text segment.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     deff
    0.75
     behö
    0.68
     devoir
    0.66
    ෲ
    0.66
     defraud
    0.66
     καθ
    0.66
     braucht
    0.66
    glut
    0.65
     fraud
    0.65
     मोहो
    0.65
    POSITIVE LOGITS
     and
    0.68
    our
    0.66
     our
    0.65
    ,<
    0.65
    рія
    0.65
    ”-
    0.63
     |.
    0.63
    SV
    0.62
    PW
    0.60
     UINT
    0.60
    Activations Density 0.000%

    No Known Activations