© 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. 15914
    Prev
    Next
    INDEX
    Explanations

    due to

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

    This neuron lights up on tokens that are entire words (i.e. full-word tokens) rather than subword fragments.

    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
    鈕
    0.79
     voltage
    0.71
     tanggung
    0.70
     spirited
    0.69
    體的
    0.69
     использоваться
    0.68
    拔
    0.68
    বন্ত
    0.67
    FRINGEMENT
    0.67
    ቀም
    0.66
    POSITIVE LOGITS
    ні
    1.04
     contado
    0.97
    та
    0.96
    də
    0.93
    да
    0.93
    ência
    0.92
     vardı
    0.91
    te
    0.90
    ا
    0.89
     verdes
    0.88
    Activations Density 0.000%

    No Known Activations