© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 417
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on sequences of uppercase letters (often with underscores) that look like constant names, configuration macros, or compile‐time flags.

    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
    -
    0.40
    ן
    0.36
    '-
    0.35
    /',
    0.34
    /",
    0.34
    rsquo
    0.33
    ‐
    0.33
    ־
    0.33
    /
    0.32
    Y
    0.32
    POSITIVE LOGITS
    ayeva
    0.38
     بجائے
    0.37
    راہ
    0.33
    ियोग
    0.32
    ərə
    0.32
    ائیں۔
    0.32
    を使用して
    0.32
     Matth
    0.32
     nameWithOwner
    0.31
    éma
    0.31
    Activations Density 0.001%

    No Known Activations