© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14678
    Prev
    Next
    INDEX
    Explanations

    GPU can / collaborate

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

    The neuron strongly activates on tokens that mark the start of list‐style entries or separators (e.g. the first word of a bullet point or the semicolon between citation entries).

    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_24_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
    tiden
    1.02
    of
    0.88
    ்க
    0.77
     Rechnung
    0.76
    듈
    0.76
     Damp
    0.75
    ры
    0.73
    tR
    0.73
    ിക്കുന്നത്
    0.71
    s
    0.71
    POSITIVE LOGITS
     appareil
    0.96
    Chorus
    0.86
     Пол
    0.82
     বাংলাদেশের
    0.79
     accompagner
    0.79
     చేరు
    0.79
     appareils
    0.76
    Ал
    0.73
    全年
    0.73
     ASSESSMENT
    0.73
    Activations Density 0.000%

    No Known Activations