© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 13791
    Prev
    Next
    INDEX
    Explanations

    If-then statements

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

    The neuron specializes in spotting major section or paragraph breaks—i.e. tokens that mark transitions such as the start of a new block (e.g. “Answer:”, “In…,” or a newline/comma at a paragraph boundary).

    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
    Ῥ
    0.80
    ricing
    0.78
     touchdowns
    0.76
     strikeouts
    0.76
    ﺮ
    0.75
     మైసూరు
    0.73
    çada
    0.73
    geld
    0.73
    figures
    0.72
    rawy
    0.72
    POSITIVE LOGITS
     institution
    0.68
     indépend
    0.68
     Upgrade
    0.67
     upgrade
    0.66
     exploratory
    0.66
     Queer
    0.66
     group
    0.65
     kai
    0.65
     Sina
    0.65
    ज
    0.65
    Activations Density 0.001%

    No Known Activations