© 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. 2584
    Prev
    Next
    INDEX
    Explanations

    ; { }

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

    The neuron strongly activates on tokens making up the word “identity” (e.g. the split pieces “ident i .”), so it’s essentially detecting occurrences of “identity.”

    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
    2
    0.74
    <0x0D>
    0.70
    <h2>
    0.68
    1
    0.68
    0
    0.65
    $
    0.64
    8
    0.64
    9
    0.63
    3
    0.62
    \
    0.61
    POSITIVE LOGITS
    〟
    0.75
    ܓ
    0.74
     💪
    0.71
     মার্কেটিং
    0.71
     ಶಿವ
    0.71
     керів
    0.70
     profitieren
    0.70
    ெடு
    0.68
     업무
    0.68
     ホイールセット
    0.67
    Activations Density 0.001%

    No Known Activations