© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 1173
    Prev
    Next
    INDEX
    Explanations

    Orangutan Moms, StartChar, I bought

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

    This neuron fires on LaTeX‐style mathematical tokens—especially Greek letter indices, braces, superscripts/subscripts, and other formula syntax.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    phar
    -0.73
    Avec
    -0.73
    berday
    -0.72
     dopo
    -0.72
    tisseur
    -0.71
    フレ
    -0.71
     utanför
    -0.71
     Rad
    -0.71
     kinderen
    -0.70
    Greg
    -0.70
    POSITIVE LOGITS
    花火
    0.81
    קי
    0.76
    ílio
    0.71
    끔
    0.71
    Medico
    0.69
    ènes
    0.69
     SUPERIOR
    0.69
    training
    0.69
    迅
    0.69
     Territor
    0.68
    Activations Density 0.008%

    No Known Activations