© 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. 7020
    Prev
    Next
    INDEX
    Explanations

    ranking lists and explanations

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

    This neuron primarily detects numeric tokens, especially four-digit numbers (e.g., years).

    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
    тен
    0.76
     Об
    0.75
    ят
    0.74
     Ш
    0.74
     зе
    0.73
     Зе
    0.72
     Кра
    0.70
    Т
    0.70
    ках
    0.70
     сен
    0.70
    POSITIVE LOGITS
    ﯼ
    0.82
    पहली
    0.80
     nếu
    0.77
    就会
    0.75
    i
    0.74
     மற்ற
    0.73
     üçün
    0.73
     benchmarks
    0.71
     chút
    0.71
     reciprocating
    0.71
    Activations Density 0.000%

    No Known Activations