© 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. 15141
    Prev
    Next
    INDEX
    Explanations

    first person statements

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

    The neuron principally fires on numeric tokens (e.g. page or time‐stamp numbers) and other document‐metadata numbers.

    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
    只会
    -1.14
     departement
    -1.05
    尚未
    -1.04
    娴
    -1.03
    オーディオ
    -0.96
     Tetrahedron
    -0.93
    确保
    -0.90
    窠
    -0.90
     оформи
    -0.90
     sacro
    -0.90
    POSITIVE LOGITS
     I
    1.54
     then
    1.41
     you
    1.39
     Então
    1.23
    chb
    1.18
    するもの
    1.16
     Obtén
    1.16
     people
    1.16
    Quindi
    1.14
     AUSSI
    1.13
    Activations Density 0.001%

    No Known Activations