© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 12171
    Prev
    Next
    INDEX
    Explanations

    This neuron consistently detects sentence- or clause-opening discourse markers and transition words (e.g. “when,” “once,” “showing,” “assignment,” “meantime”)—the first content tokens that introduce a new clause or sentence.

    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
     permainan
    0.98
     juegos
    0.86
     Juegos
    0.83
     fuerza
    0.80
     bởi
    0.79
     réduit
    0.79
     humide
    0.77
    实
    0.75
     dầu
    0.74
     ajoute
    0.74
    POSITIVE LOGITS
    ibo
    0.82
    ao
    0.75
    RootNamespace
    0.70
    PubMed
    0.68
    ijiet
    0.68
    ophil
    0.68
     correlates
    0.67
    चंद्र
    0.66
    ip
    0.66
    ст
    0.65
    Activations Density 0.145%

    No Known Activations