© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 130990
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on standalone numeric tokens (integer literals) in the code/comment.

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

    documentation comments and code

    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
     both
    -1.66
     even
    -1.63
     from
    -1.60
     if
    -1.47
     which
    -1.42
     that
    -1.39
     its
    -1.38
     regardless
    -1.38
     for
    -1.37
     those
    -1.33
    POSITIVE LOGITS
    does
    1.21
    Y
    1.15
     wondrous
    1.14
    M
    1.14
    N
    1.10
     ß
    1.07
    Aa
    1.05
     AMAZING
    1.05
    ED
    1.05
     interessante
    1.04
    Activations Density 0.002%

    No Known Activations