© 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. 7326
    Prev
    Next
    INDEX
    Explanations

    rising through or above

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

    The neuron strongly activates on the substring “aphor” (as in “Metaphor”), i.e. it detects occurrences of the word “metaphor.”

    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
     toxicants
    0.87
     convolutions
    0.84
    дцать
    0.83
    граммы
    0.82
     каме
    0.79
    𝔱
    0.78
    annotations
    0.77
    馗
    0.77
     annotations
    0.76
     продукции
    0.76
    POSITIVE LOGITS
    jár
    0.87
     Etter
    0.84
    ava
    0.78
     Прави
    0.76
    важа
    0.73
    ählen
    0.73
     Truy
    0.72
    ilty
    0.71
    aya
    0.70
    jangan
    0.70
    Activations Density 0.000%

    No Known Activations