© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 11687
    Prev
    Next
    INDEX
    Explanations

    style, pattern, finish

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

    The neuron fires strongly on capitalized attribute labels (like “Type,” “Age range,” “Material,” “Pattern,” “Occasion,” etc.) in product specification sections of ecommerce listings.

    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
    Trong
    0.65
    Với
    0.64
     Crater
    0.64
    workspaceFolder
    0.63
     Với
    0.63
    Puede
    0.63
    㝌
    0.63
    ᗜ
    0.63
     ด้วย
    0.63
    راً
    0.62
    POSITIVE LOGITS
     conversely
    0.64
     b
    0.63
     resigned
    0.63
     raped
    0.61
     spoke
    0.61
     cut
    0.60
     said
    0.59
     slash
    0.59
    0.59
     alia
    0.58
    Activations Density 0.001%

    No Known Activations