© 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. 4156
    Prev
    Next
    INDEX
    Explanations

    4

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

    The neuron strongly activates on standalone numerals (and their trailing punctuation) used as list or step markers.

    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
    aj
    0.75
    {
    0.73
    ج
    0.73
    orr
    0.71
    ill
    0.70
    ak
    0.68
    ول
    0.67
    து
    0.66
    yr
    0.66
    ay
    0.65
    POSITIVE LOGITS
     Doesn
    0.66
     Corruption
    0.61
     परिश्रम
    0.59
    ۔
    0.59
     Ombudsman
    0.59
     Schuld
    0.58
     दिखी
    0.57
     indemnify
    0.57
     documenti
    0.57
     mémoire
    0.57
    Activations Density 0.100%

    No Known Activations