© 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. 6497
    Prev
    Next
    INDEX
    Explanations

    ad- prefix

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

    The neuron spots tokens beginning with the substring “ad,” such as those marking ads or ad-related strings.

    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
     നടത്തി
    0.53
     державної
    0.50
    เง
    0.50
    众人
    0.49
    depth
    0.49
    insect
    0.49
    wydd
    0.48
     来
    0.48
     माध्यमा
    0.48
    angel
    0.48
    POSITIVE LOGITS
     Ad
    0.89
     Ад
    0.88
    elaide
    0.86
     アド
    0.78
     hoc
    0.75
     adh
    0.73
    olphe
    0.73
    Ad
    0.72
     آد
    0.70
     adi
    0.68
    Activations Density 0.177%

    No Known Activations