© 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. 10-GEMMASCOPE-RES-131K
    7. 69809
    Prev
    Next
    INDEX
    Explanations

    surprise or shock

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

    The neuron detects occurrences of expressions of surprise or unexpectedness (e.g. “surprised”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     Presumably
    -1.73
     to
    -1.73
    逦
    -1.66
     a
    -1.59
     vermutlich
    -1.55
     Also
    -1.52
     circumvent
    -1.52
     Furthermore
    -1.51
     Moreover
    -1.50
    ࿚
    -1.50
    POSITIVE LOGITS
     there
    1.73
     that
    1.50
    foundland
    1.48
    s
    1.45
     totalmente
    1.38
    合适
    1.38
    ՚
    1.37
     kies
    1.34
     stap
    1.34
    </b>
    1.33
    Activations Density 0.015%

    No Known Activations