© 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. 1060
    Prev
    Next
    INDEX
    Explanations

    This neuron fires most strongly on unusually large or off-distribution token embeddings—i.e. rare, garbled, or “noisy” 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
     ></
    0.73
     :::
    0.70
     Issled
    0.68
     эксплуатации
    0.66
     вещества
    0.65
    педи
    0.64
     веществ
    0.64
    বো
    0.63
    也就
    0.63
    igkeiten
    0.63
    POSITIVE LOGITS
     siren
    0.84
    TMP
    0.79
    てる
    0.78
     vitality
    0.73
     kendisine
    0.73
     voulez
    0.72
    y
    0.72
     pert
    0.71
     집
    0.71
     doit
    0.71
    Activations Density 0.000%

    No Known Activations