© 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. 130469
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on numeric tokens (individual numbers or numeric literals).

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

    "Is composite number" or "nearest integer"

    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
    iong
    -1.07
     in
    -1.03
    XE
    -0.96
    ԝ
    -0.95
    thè
    -0.91
    痔
    -0.88
    зь
    -0.88
    ква
    -0.88
    кла
    -0.88
    amento
    -0.88
    POSITIVE LOGITS
     dinheiro
    1.13
    wesenheit
    1.09
    Charakter
    1.06
     fewer
    1.03
    だったが
    1.02
     ausreichend
    1.02
    Recept
    1.01
    だった
    0.97
    やった
    0.96
    Pendant
    0.94
    Activations Density 0.012%

    No Known Activations