© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 8952
    Prev
    Next
    INDEX
    Explanations

    whether you're

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

    The neuron fires strongly on tokens that refer back to an earlier subject or signal a narrative shift—e.g. pronouns like “they” and connective adverbs like “then.”

    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
     లేదా
    0.63
     ktoré
    0.62
    ที่ไม่
    0.60
     které
    0.59
    を取得
    0.58
     أو
    0.57
     která
    0.56
    ﺑ
    0.56
     えっ
    0.55
     اتارنا
    0.55
    POSITIVE LOGITS
    ీ
    0.73
    0.69
    न
    0.66
    .
    0.61
    ק
    0.58
    на
    0.56
    ل
    0.55
     suff
    0.54
    ா
    0.54
    al
    0.53
    Activations Density 0.014%

    No Known Activations