© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 15858
    Prev
    Next
    INDEX
    Explanations

    people in studies or services

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

    The neuron activates on numeric tokens or sequences (digits, counts, percentages, years, etc.), i.e. it finds numbers or numeric expressions.

    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
     our
    -1.33
    ẩ
    -1.01
     we
    -1.00
     when
    -0.99
    Our
    -0.98
     while
    -0.96
     disparos
    -0.94
     its
    -0.93
    ковые
    -0.93
     JUST
    -0.91
    POSITIVE LOGITS
     either
    1.71
     möjlighet
    1.34
     can
    1.24
    もありますが
    1.21
    спомина
    1.20
    either
    1.18
     receive
    1.12
     receber
    1.10
     alltså
    1.10
    rance
    1.08
    Activations Density 0.053%

    No Known Activations