© 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. 31-GEMMASCOPE-2-RES-16K
    7. 5607
    Prev
    Next
    INDEX
    Explanations

    citrus fruits

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

    The neuron strongly activates on ingredient and flavor component terms (especially citrus fruits, oils, and related chemical additives).

    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
    mila
    0.63
    鞍
    0.62
    蚝
    0.59
    瞑
    0.56
    駝
    0.54
    াম
    0.54
    驼
    0.54
    TabLayout
    0.53
    GRAY
    0.53
    打ち
    0.52
    POSITIVE LOGITS
     citrus
    2.06
     Citrus
    2.00
    🍊
    1.93
    🍋
    1.92
     lemons
    1.91
     lemon
    1.81
     oranges
    1.80
     Lemon
    1.60
    Lemon
    1.60
    lemon
    1.59
    Activations Density 0.291%

    No Known Activations