© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 2867
    Prev
    Next
    INDEX
    Explanations

    entity followed by context

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

    The neuron consistently turns on for uncommon, multi­-syllable nouns and proper names—i.e. named entities like “tamarin,” “fishtail palm,” “Phobos,” “Karl­sborg Fortress,” “Sadako,” “calendar,” “ball,” etc.

    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
    /
    0.82
     গুলোতে
    0.66
     tersebut
    0.66
    !!!
    0.64
     গুলো
    0.64
     maupun
    0.63
    !!
    0.61
     escalar
    0.61
     }}$.
    0.59
    非常的
    0.59
    POSITIVE LOGITS
     roundtable
    0.76
    abortion
    0.73
    ʐ
    0.72
     上海
    0.71
    博物馆
    0.70
    댓글
    0.69
    ከሰ
    0.68
     అభిమా
    0.68
     музей
    0.67
     έγινε
    0.66
    Activations Density 0.175%

    No Known Activations