© 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. 22-GEMMASCOPE-RES-131K
    7. 130069
    Prev
    Next
    INDEX
    Explanations

    The neuron is specially keyed to spotting numeric literals (number constants) in code snippets.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    code syntax and programming constructs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    ⟹
    -0.98
    wist
    -0.89
    驹
    -0.88
    waf
    -0.87
    isez
    -0.87
    &
    -0.84
    balon
    -0.84
     kú
    -0.84
     kateg
    -0.84
    ização
    -0.80
    POSITIVE LOGITS
     radic
    0.76
    可能的
    0.75
     CALC
    0.75
    私たちの
    0.74
     tortilla
    0.73
     Ripple
    0.73
    Opere
    0.72
     FED
    0.72
    线条
    0.72
     駅
    0.71
    Activations Density 0.019%

    No Known Activations