© 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. 31-GEMMASCOPE-2-RES-16K
    7. 6451
    Prev
    Next
    INDEX
    Explanations

    Spelling corrections and phrase replacements

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

    The neuron fires on numeric tokens—digits, numbers (including decimals), and other numeral‐like strings 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_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
    рное
    0.55
    Pedido
    0.54
    elseif
    0.52
    Tür
    0.52
    胶
    0.51
    else
    0.51
    FileSync
    0.51
    C
    0.50
    Callback
    0.50
    *_
    0.50
    POSITIVE LOGITS
     whos
    0.62
     ?.
    0.61
     brings
    0.61
     Punjabi
    0.61
     yakni
    0.60
     entails
    0.57
     ?"
    0.57
     "*.
    0.57
     ២
    0.57
     ?”
    0.56
    Activations Density 0.038%

    No Known Activations