© 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. 113900
    Prev
    Next
    INDEX
    Explanations

    watching something happen

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

    The neuron flags instances of observing or viewing events, especially occurrences of “watch” (and its past tense “watched”) often in “watch…as” or “watched…with” contexts.

    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
    牁
    -1.88
     decorada
    -1.59
     you
    -1.59
     decoradas
    -1.58
     equipada
    -1.58
     your
    -1.56
    縠
    -1.54
     rápidas
    -1.54
    rrggbb
    -1.53
     dispuestos
    -1.49
    POSITIVE LOGITS
    8
    2.08
     for
    1.67
    would
    1.61
    \
    1.52
     "
    1.51
    ].
    1.50
    9
    1.48
     with
    1.48
    .'
    1.45
     others
    1.44
    Activations Density 0.009%

    No Known Activations