© 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. 24-GEMMASCOPE-2-RES-16K
    7. 9935
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively fires on tokens involved in licensing, attribution, or selling—i.e. words from boilerplate copyright/“for sale” sections.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    extremely high activation values that appear to be anomalies or errors in the data, often occurring near sentence or document boundaries.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    terms associated with publishing and intellectual property/commercial transactions (e.g., authorship, licensing, selling, applicability).

    oai_token-act-pair · gpt-5Triggered 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
    el
    0.77
    数据
    0.77
    ॉज
    0.76
    ১৬
    0.75
    图
    0.74
    as
    0.71
    三星
    0.70
    Scenic
    0.70
    طق
    0.69
    Hayes
    0.69
    POSITIVE LOGITS
     пути
    0.80
     neuest
    0.73
     produkter
    0.70
    десят
    0.69
     dame
    0.68
    ור
    0.68
     mercantil
    0.67
    "+
    0.66
    iskt
    0.66
     hoeveel
    0.66
    Activations Density 0.000%

    No Known Activations