© 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. 31-GEMMASCOPE-2-RES-16K
    7. 12570
    Prev
    Next
    INDEX
    Explanations

    documentation comments

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

    This neuron activates on comment delimiters and related punctuation markers in code (e.g., /*, */, //, ///, @, and other symbols used to mark or format comments).

    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_31_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
    mData
    0.63
    x
    0.61
    ti
    0.60
    fail
    0.59
    k
    0.57
    cf
    0.57
    ch
    0.57
    hi
    0.56
    was
    0.55
    xf
    0.55
    POSITIVE LOGITS
     使っ
    0.61
    )\}
    0.61
     constants
    0.59
     ज्योति
    0.57
     영역
    0.56
     VARIABLES
    0.56
     হেল
    0.55
     Gemeins
    0.55
    ității
    0.55
     분야
    0.55
    Activations Density 0.001%

    No Known Activations