© 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. 31-GEMMASCOPE-2-RES-16K
    7. 9417
    Prev
    Next
    INDEX
    Explanations

    developed by

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

    This neuron strongly activates on the preposition “by” when used to mark agency or attribution (e.g. “funded by,” “sold by,” “transcribed by”).

    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
    ক
    0.83
     ഇത്
    0.77
    的作用
    0.71
    에
    0.71
    기
    0.71
    BASE
    0.71
    que
    0.70
     여기
    0.70
     ประเภท
    0.70
    자
    0.70
    POSITIVE LOGITS
    zantine
    1.59
     virtue
    1.34
     means
    1.08
    衷
    1.01
     candlelight
    0.98
     dint
    0.97
    zant
    0.95
     mistake
    0.94
     the
    0.92
     către
    0.91
    Activations Density 0.253%

    No Known Activations