© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 83412
    Prev
    Next
    INDEX
    Explanations

    fold cross validation

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

    This neuron is strongly activated by programming identifiers—especially capitalized class or function names, library/module names, and other code tokens in Python source.

    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
     گیل
    -0.71
    ätzt
    -0.71
    谚
    -0.68
     Pfer
    -0.67
     horas
    -0.66
    acters
    -0.66
     Fett
    -0.66
    aling
    -0.65
     plat
    -0.64
     المنام
    -0.64
    POSITIVE LOGITS
     folds
    2.31
     fold
    2.02
    fold
    1.92
    folds
    1.91
    Fold
    1.88
     cross
    1.80
     Fold
    1.69
    FOLD
    1.64
    cross
    1.45
     folding
    1.40
    Activations Density 0.017%

    No Known Activations