© 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. 65682
    Prev
    Next
    INDEX
    Explanations

    names and compound words

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

    The neuron responds to uncommon or domain‐specific tokens—especially proper names and specialized vocabulary—by strongly activating on those rare terms.

    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
     this
    -1.30
     any
    -1.24
     what
    -1.17
     one
    -1.15
     in
    -1.12
     we
    -1.10
     these
    -1.07
     with
    -1.04
    ”，
    -1.03
     but
    -1.02
    POSITIVE LOGITS
     procedimiento
    1.30
     actuales
    1.11
    lccn
    1.05
     Detalles
    1.04
    ויים
    1.03
     bohater
    1.02
    jango
    1.00
     izquier
    1.00
    mband
    0.99
    кумулятор
    0.98
    Activations Density 0.195%

    No Known Activations