© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 107958
    Prev
    Next
    INDEX
    Explanations

    file annotation tropics assembly function gold

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

    The neuron fires on structural metadata tokens in code or markup—things like file‐related identifiers (“file”), XML/HTML tag names, and the assignment operator (“=”).

    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
     fär
    -0.78
    portug
    -0.71
     mía
    -0.70
     ním
    -0.69
     POV
    -0.69
     Colors
    -0.68
     Поли
    -0.68
    nuss
    -0.67
     материалы
    -0.66
    νή
    -0.66
    POSITIVE LOGITS
    ль
    0.65
    vokat
    0.64
    villa
    0.63
    listi
    0.63
    しかし
    0.63
    outputFile
    0.61
    点点
    0.61
    reducers
    0.61
    ísticas
    0.59
    berto
    0.59
    Activations Density 0.057%

    No Known Activations