© 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. 6617
    Prev
    Next
    INDEX
    Explanations

    in, by, below, with

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

    The neuron strongly activates on numerical tokens—especially floating‐point numbers (decimal values) 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
     ultra
    0.71
    ultra
    0.68
    สำหรับการ
    0.67
     ult
    0.62
    ायला
    0.62
     same
    0.60
    4
    0.60
     ultr
    0.59
    同样的
    0.59
    msız
    0.59
    POSITIVE LOGITS
     manifested
    0.70
     笑
    0.66
     здесь
    0.65
     प्रीवियस
    0.65
     அவர்களிடம்
    0.64
     своих
    0.63
    aruhi
    0.62
    茟
    0.62
     यहाँ
    0.61
    স্ফ
    0.61
    Activations Density 0.088%

    No Known Activations