© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 6124
    Prev
    Next
    INDEX
    Explanations

    Academic, scientific, definition, data, survival

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

    This neuron fires on the very first token of a new paragraph or section heading—that is, it detects paragraph/section boundaries by spotting the initial word of each block.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    commas that separate items in lists or follow introductory phrases.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    mentions of nouns, especially concrete places, objects, or named entities.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_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
    2
    0.83
    1
    0.78
    3
    0.77
    4
    0.74
    5
    0.73
    好多
    0.73
    бы
    0.70
    9
    0.70
    Workbook
    0.69
    щается
    0.67
    POSITIVE LOGITS
     située
    0.89
     lembut
    0.82
     tư
    0.80
     scientifique
    0.79
    ⤠
    0.78
     консу
    0.77
     tertinggi
    0.77
     situada
    0.76
     сти
    0.75
     сама
    0.75
    Activations Density 0.000%

    No Known Activations