© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 12461
    Prev
    Next
    INDEX
    Explanations

    behold and variations

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

    The neuron strongly activates on the archaic exclamation “behold” (including in phrases like “lo and behold”).

    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
    i
    -3.28
    x
    -2.97
    b
    -2.47
     After
    -2.31
    "
    -2.30
    g
    -2.25
    ka
    -2.23
    van
    -2.23
     regal
    -2.22
    !"
    -2.19
    POSITIVE LOGITS
    寕
    3.69
    珝
    2.84
    鲧
    2.73
    比如说
    2.72
     花纹
    2.72
    ℟
    2.70
    禤
    2.69
    因为它
    2.66
    錞
    2.64
    壼
    2.59
    Activations Density 0.002%

    No Known Activations