© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1071
    Prev
    Next
    INDEX
    Explanations

    intended destination or state

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

    The neuron primarily fires on the definite article “the” (and other common function words) when they appear in formal, legal‐ or policy‐style boilerplate text.

    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_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
     the
    0.73
     dây
    0.71
    illary
    0.71
    ఞ
    0.69
     and
    0.68
    ámenes
    0.67
    urile
    0.66
     图片
    0.66
    nThe
    0.65
    ⬝
    0.65
    POSITIVE LOGITS
     following
    1.27
     undersigned
    1.23
     appropriate
    1.13
     siguientes
    1.13
     respective
    1.10
    following
    1.08
     aforementioned
    1.04
     entsprechenden
    1.04
     jeweiligen
    1.03
     corresponding
    1.02
    Activations Density 0.679%

    No Known Activations