© 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. 8787
    Prev
    Next
    INDEX
    Explanations

    end of sentence punctuation

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

    The neuron strongly activates on product‐listing phrases that combine a retailer name (e.g. “Walmart”), a color or descriptor (e.g. “white”), and a product (e.g. “baby cribs”).

    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
     evas
    0.75
     (&
    0.71
     (!)
    0.71
    é
    0.70
     Junio
    0.68
     (
    0.66
     Nonlinear
    0.65
    և
    0.65
     entrenched
    0.63
    োম
    0.63
    POSITIVE LOGITS
     الذي
    0.75
     radiographs
    0.74
    ऱ्यावर
    0.73
     তৈরির
    0.66
    anzi
    0.66
     માં
    0.65
     radi
    0.64
     შესახებ
    0.64
    radi
    0.64
     radiometric
    0.63
    Activations Density 0.094%

    No Known Activations