© 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. 2315
    Prev
    Next
    INDEX
    Explanations

    text following "learned"

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

    The neuron fires strongly at the start or end of self‐contained text blocks (i.e. it marks segment boundaries).

    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
     hapless
    0.72
     FRANCE
    0.70
     Chines
    0.68
     ö
    0.67
     Dwar
    0.67
     INTO
    0.66
    Prin
    0.66
     UPPER
    0.66
     Bri
    0.66
     Prince
    0.65
    POSITIVE LOGITS
     hatchback
    0.89
    вая
    0.86
     связано
    0.85
    碰
    0.84
    вые
    0.84
    гл
    0.84
    мы
    0.83
    ση
    0.80
     связана
    0.80
    ົບ
    0.79
    Activations Density 0.000%

    No Known Activations