© 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. 130359
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric literals (tokens composed of digit sequences) in code.

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

    code blocks and namespaces

    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
     mettant
    -0.86
     dasselbe
    -0.85
     estudian
    -0.83
     cuentan
    -0.79
     other
    -0.79
     olvido
    -0.78
     neglecting
    -0.77
    다고
    -0.77
     barista
    -0.77
     parecen
    -0.77
    POSITIVE LOGITS
     provoque
    0.88
     '../../../../
    0.84
     Freien
    0.77
    jski
    0.77
    mgmt
    0.76
    hộp
    0.76
     malade
    0.75
    ]{\
    0.75
     autorisation
    0.75
    ㅏ
    0.75
    Activations Density 0.025%

    No Known Activations