© 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. 86521
    Prev
    Next
    INDEX
    Explanations

    expressing surprise or inquiry

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

    The neuron detects speaker-stance or opinion phrases—small discourse markers expressing surprise or personal commentary (e.g. “no wonder,” “I can’t believe,” “I bet,” “how come”).

    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
     and
    -1.80
     bekannten
    -1.80
     in
    -1.80
     This
    -1.80
    The
    -1.78
    _
    -1.73
    PENUTUP
    -1.70
     إذا
    -1.65
    """
    -1.62
    １
    -1.60
    POSITIVE LOGITS
     wrze
    1.58
    埽
    1.57
     garante
    1.52
    泚
    1.51
     lesquelles
    1.49
     metic
    1.48
    iczne
    1.48
     them
    1.48
    ्रा
    1.48
     dispo
    1.46
    Activations Density 0.027%

    No Known Activations