© 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. 10447
    Prev
    Next
    INDEX
    Explanations

    deep neural networks

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

    The neuron strongly activates on uncommon, content‐bearing tokens—key nouns and verbs that carry semantic weight (e.g. “decision,” “advocate,” “workshop,” “prisoner,” “memories,” “Ford”).

    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
    ETT
    0.79
    ☘
    0.77
    OMX
    0.75
    QUIS
    0.75
     Remarks
    0.73
    доб
    0.72
    例如
    0.71
     וא
    0.71
    лика
    0.70
    сту
    0.70
    POSITIVE LOGITS
     jogar
    0.96
     estadounidense
    0.94
     automática
    0.91
     sufrir
    0.91
    ซ
    0.91
     jogo
    0.90
     manteniendo
    0.89
     procedente
    0.89
     juta
    0.88
     penumpang
    0.87
    Activations Density 0.000%

    No Known Activations