© 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. 938
    Prev
    Next
    INDEX
    Explanations

    structured explanations

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

    The neuron is acting as a “common‐word” detector, firing strongly on very high‐frequency function words (e.g. “of,” “the,” “both,” “day,” etc.).

    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
    ிரிய
    0.89
    ek
    0.77
    ekten
    0.74
    ният
    0.74
    eit
    0.74
    enskap
    0.73
    equity
    0.71
    貂
    0.71
    惀
    0.71
    رم
    0.70
    POSITIVE LOGITS
    이랑
    0.73
    be
    0.69
     Є
    0.69
     मगर
    0.68
     Ос
    0.67
     Jalan
    0.66
     C
    0.66
     ,
    0.64
     È
    0.63
    ც
    0.63
    Activations Density 0.627%

    No Known Activations