© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 15501
    Prev
    Next
    INDEX
    Explanations

    attracting attention

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

    The neuron activates on occurrences of the word “attention” (and close variants) in the text.

    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
    The
    -2.09
    这些
    -2.03
    ſelves
    -1.89
    Already
    -1.73
     to
    -1.71
    Coming
    -1.66
    懨
    -1.65
    (
    -1.65
    牁
    -1.60
    What
    -1.59
    POSITIVE LOGITS
     incidentally
    1.43
    苼
    1.41
     about
    1.38
    lessly
    1.36
    d
    1.34
    ，，
    1.31
    n
    1.31
     иной
    1.29
    跸
    1.28
    眛
    1.28
    Activations Density 0.010%

    No Known Activations