© 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. 129727
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on math expressions where a single variable is being set equal to a constant (e.g. “y = 1”).

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

    numbers near $\pm$

    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
    био
    -0.96
     EQU
    -0.92
     EventHandler
    -0.88
    Attribute
    -0.85
    forName
    -0.84
    ‌ترین
    -0.84
     детских
    -0.84
     домашних
    -0.84
     étro
    -0.81
    кри
    -0.80
    POSITIVE LOGITS
     and
    0.96
     Тогда
    0.94
     then
    0.88
    旼
    0.86
     dises
    0.85
    :[{
    0.83
     этом
    0.79
     margar
    0.79
    жала
    0.79
    Significado
    0.78
    Activations Density 0.026%

    No Known Activations