© 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. 1522
    Prev
    Next
    INDEX
    Explanations

    std::endl

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

    The neuron fires on punctuation‐heavy code tokens—especially backslashes, escape sequences (like “\n”), and stream‐insertion operators (like “<<”)—rather than on normal words.

    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
    ↵
    0.63
    િ
    0.40
    Sidebar
    0.40
    LeftSide
    0.39
    othing
    0.39
    Notation
    0.39
    Ellipse
    0.38
     if
    0.38
    HairColor
    0.38
     lakukan
    0.38
    POSITIVE LOGITS
    oeste
    0.49
    ॕ
    0.48
    d
    0.48
    вая
    0.47
    ship
    0.45
     стен
    0.45
    iniai
    0.45
    gène
    0.45
    ά
    0.45
    ísticas
    0.44
    Activations Density 0.000%

    No Known Activations