© 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. 8998
    Prev
    Next
    INDEX
    Explanations

    technical research and analysis

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

    The neuron fires strongly on the very first content word at the start of each new text segment (i.e. it marks segment-initial tokens).

    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
     humains
    0.88
    ം
    0.86
     fysis
    0.82
     hennes
    0.80
    reibt
    0.78
     individus
    0.78
    𝐡
    0.78
    𝐮
    0.77
     lancéolées
    0.75
     spaceShip
    0.74
    POSITIVE LOGITS
    ن
    0.83
     Translational
    0.72
     Gateway
    0.70
    א
    0.70
     בית
    0.69
    седа
    0.68
    oric
    0.68
    orical
    0.68
     Meteorological
    0.67
    样子
    0.67
    Activations Density 0.000%

    No Known Activations