© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 5866
    Prev
    Next
    INDEX
    Explanations

    No X found

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

    The neuron strongly responds to numerals and other list‐ or quantity‐marking tokens (e.g. “0,” “first,” “two,” “month,” etc.), i.e. it detects explicit numeric or enumerative expressions.

    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
     è
    0.92
    可以
    0.90
     Tolst
    0.80
     può
    0.77
    "
    0.75
    可以是
    0.74
    ”，
    0.72
    ”
    0.72
    慘
    0.70
     são
    0.69
    POSITIVE LOGITS
    रिडोर
    0.98
    ządz
    0.82
    чный
    0.80
    менный
    0.80
     namesake
    0.80
    товый
    0.80
    лэ
    0.79
    まるで
    0.79
    ्यूर
    0.78
     grapevine
    0.76
    Activations Density 0.000%

    No Known Activations