© 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. 20670
    Prev
    Next
    INDEX
    Explanations

    surprise, shock, secret, or unexpected reveal

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

    The neuron fires on emphatic or ironic discourse markers (e.g. “Shocker,” “Surprise:,” “funny enough,” “behold”) used to flag sarcasm or dramatic emphasis.

    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
     its
    -1.29
     jego
    -1.27
    <bos>
    -1.27
     его
    -1.25
    ImageField
    -1.21
     helft
    -1.18
    MAXN
    -1.18
    |()
    -1.14
    greenrobot
    -1.13
     kobieta
    -1.13
    POSITIVE LOGITS
     gerçekten
    1.40
     other
    1.27
    Use
    1.16
    Other
    1.16
    気で
    1.15
     çünkü
    1.14
     postre
    1.13
     кое
    1.12
    Creating
    1.12
    其他的
    1.09
    Activations Density 0.047%

    No Known Activations