© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 59005
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and operations

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

    The neuron strongly activates on the question‐mark token “?”, i.e. it detects question endings.

    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
    </sup>
    -2.00
     Inſ
    -1.66
    鴕
    -1.66
    }{
    -1.60
     喜欢
    -1.59
    ︲
    -1.58
     erfolgreichen
    -1.57
    }=\
    -1.55
    接下來
    -1.55
     玛
    -1.51
    POSITIVE LOGITS
    ?
    3.00
    ’
    2.91
    "$_
    2.19
    ↵
    2.09
     Its
    2.06
    +"'
    1.85
     in
    1.84
      
    1.80
    {
    1.80
    “
    1.77
    Activations Density 0.142%

    No Known Activations