© 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. 129792
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on numeric tokens (sequences of digits), i.e. it detects numbers.

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

    java code 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
     trám
    -1.13
     akty
    -1.05
     kolek
    -1.05
     komer
    -1.01
     lác
    -1.00
     saurait
    -0.98
    mø
    -0.98
    łania
    -0.98
     tunik
    -0.96
     instruk
    -0.96
    POSITIVE LOGITS
     first
    0.95
     also
    0.91
     over
    0.89
    地说
    0.88
    Consejo
    0.84
     has
    0.82
     three
    0.82
     high
    0.81
     predomin
    0.81
     helps
    0.79
    Activations Density 0.046%

    No Known Activations