© 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. 36658
    Prev
    Next
    INDEX
    Explanations

    code punctuation and programming terms

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

    The neuron strongly activates on tokens that are numeric literals (tokens containing digits, including integers and decimals).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    Vá
    -1.05
     expuesto
    -1.02
    ベトナム
    -0.97
     rám
    -0.93
     vertik
    -0.93
    giques
    -0.92
     queijo
    -0.92
    INCENT
    -0.91
    ྅
    -0.91
    GRAPHS
    -0.90
    POSITIVE LOGITS
    ордина
    1.09
     such
    0.95
     found
    0.94
     themselves
    0.93
    报记者
    0.87
     if
    0.84
     *
    0.82
     If
    0.81
    rak
    0.81
    使って
    0.81
    Activations Density 0.035%

    No Known Activations