© 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. 10726
    Prev
    Next
    INDEX
    Explanations

    code blocks

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

    The neuron detects discourse-level connectors or adverbial transition words (e.g. “actually,” “as a result,” “when,” “further,” “hence,” “ends up”) that signal causal or sequential relations.

    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
    liques
    0.85
    avkh
    0.83
     少し
    0.81
    臆
    0.80
    ्लाई
    0.79
    rawdę
    0.79
    fromj
    0.77
     Это
    0.77
    णाला
    0.77
     anisotropy
    0.76
    POSITIVE LOGITS
     &
    0.79
    /
    0.74
     and
    0.71
     hoặc
    0.70
     or
    0.67
     oder
    0.67
     C
    0.66
     và
    0.65
     mainstream
    0.64
     merupakan
    0.64
    Activations Density 0.092%

    No Known Activations