© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25198
    Prev
    Next
    INDEX
    Explanations

    questions and instructions

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

    This neuron spotlights the structural markers of problem statements—e.g. numbered or lettered part labels (like “(1)”, “(i)”, “(a)”, etc.) and associated instruction words (“Find”, “Show”, “lim”, etc.).

    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
     also
    -1.21
     gewünschten
    -1.11
    jaket
    -1.10
    :</
    -1.07
    caucasian
    -1.05
     that
    -1.04
     Aufenthalt
    -1.02
    🩱
    -0.98
    gigi
    -0.98
    ductos
    -0.94
    POSITIVE LOGITS
     zunächst
    1.30
    ga
    1.01
    峻
    0.97
    web
    0.96
    is
    0.96
     通販
    0.94
    set
    0.94
    DTD
    0.94
     hinauf
    0.93
    🤢
    0.93
    Activations Density 0.015%

    No Known Activations