© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 8381
    Prev
    Next
    INDEX
    Explanations

    numbers and code structures

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

    The neuron strongly activates on content‐bearing (non–stop) words—i.e. longer nouns, verbs, and adjectives—while ignoring common function words.

    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
     
    0.84
     (
    0.82
     .
    0.75
     =
    0.71
     {
    0.69
     [
    0.65
     \
    0.64
     see
    0.64
    in
    0.63
    id
    0.63
    POSITIVE LOGITS
     volontà
    0.83
    䣫
    0.83
     retornar
    0.77
     voluntad
    0.75
    䀼
    0.75
    profitLoss
    0.73
    伍章
    0.72
    酤
    0.71
     returnPacket
    0.71
     trueMap
    0.71
    Activations Density 0.058%

    No Known Activations