© 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. 5395
    Prev
    Next
    INDEX
    Explanations

    starting lists

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

    The neuron strongly responds to very high‐frequency function and formatting words (like “of,” “the,” “description,” “more,” etc.) that are typical in boilerplate legal or policy‐style text.

    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
    aderamente
    0.79
    ్ఞ
    0.73
    nicheskij
    0.70
    tochy
    0.69
    ucristo
    0.69
     بودم
    0.69
    ukiyoe
    0.68
    Ᏺ
    0.68
    ாதாரண
    0.68
    
    0.68
    POSITIVE LOGITS
    こちらも
    1.19
     again
    1.15
     ebenfalls
    1.13
     Again
    0.98
     opět
    0.96
    これも
    0.90
    Again
    0.88
     also
    0.87
    again
    0.86
     AGAIN
    0.86
    Activations Density 1.786%

    No Known Activations