© 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. 24-GEMMASCOPE-2-RES-16K
    7. 4043
    Prev
    Next
    INDEX
    Explanations

    repeatedly steps through

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

    The neuron strongly activates on numeric tokens—years, dates, measurements, or other number sequences in the 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_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
     ר
    0.80
     ת
    0.77
    ión
    0.76
     mos
    0.72
     근데
    0.71
     wre
    0.71
     נו
    0.70
    য়ন
    0.68
     người
    0.67
     prioritized
    0.66
    POSITIVE LOGITS
    Carta
    0.88
    的一
    0.82
    比例
    0.80
    灸
    0.79
     klicken
    0.78
    Го
    0.76
    Bonus
    0.75
     zweite
    0.74
    Ви
    0.73
    したら
    0.73
    Activations Density 0.000%

    No Known Activations