© 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. 22-GEMMASCOPE-RES-131K
    7. 130393
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on mentions of perfect-power terminology (e.g. “square,” “cube,” “fourth power,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    whereas, while, or contrast

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    ,
    -1.15
     olmayan
    -0.75
    :
    -0.74
    5
    -0.74
    なので
    -0.74
     to
    -0.73
     assumptions
    -0.72
     —
    -0.72
     so
    -0.70
    zusetzen
    -0.70
    POSITIVE LOGITS
     whereas
    1.75
     tandis
    1.62
    whereas
    1.29
     ส่วน
    1.27
     Whereas
    1.25
     mientras
    1.20
    」、
    1.17
    Whereas
    1.17
     natomiast
    1.16
     sedangkan
    1.14
    Activations Density 0.272%

    No Known Activations