© 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. 15208
    Prev
    Next
    INDEX
    Explanations

    code structure indexing

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

    The neuron fires on the first non-function (content) word that appears immediately after a major punctuation break (colon, semicolon, parentheses, etc.).

    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
    ляет
    0.79
    儡
    0.79
    ählt
    0.78
    coprote
    0.78
     accad
    0.77
    cü
    0.76
     Leid
    0.76
    openjdk
    0.76
     afectan
    0.75
     ఘటన
    0.75
    POSITIVE LOGITS
    er
    0.72
    ר
    0.69
    🍾
    0.68
    intosh
    0.68
    গাঁ
    0.65
    MESH
    0.64
     fittings
    0.63
    న్
    0.62
     intentions
    0.62
     организова
    0.62
    Activations Density 0.000%

    No Known Activations