© 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. 1247
    Prev
    Next
    INDEX
    Explanations

    organized in layers. font

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

    The neuron strongly responds to tokens that are uppercase acronyms, model names, or alphanumeric technical identifiers (e.g., “US,” “IVUS,” “2D,” “3D,” “VGG,” etc.).

    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
    Magento
    0.57
    ()">
    0.51
    CURSOR
    0.50
    Timestamp
    0.50
    CHUNK
    0.49
    ◞
    0.49
    ۰
    0.48
     '>
    0.47
    =',
    0.46
    TIMESTAMP
    0.46
    POSITIVE LOGITS
    的存在
    0.46
     neighbor
    0.44
    леты
    0.43
    一座
    0.41
    으로부터
    0.41
    に取り組
    0.40
     यही
    0.39
    urb
    0.39
    and
    0.38
     harbor
    0.38
    Activations Density 0.181%

    No Known Activations