© 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. 22-GEMMASCOPE-RES-131K
    7. 24965
    Prev
    Next
    INDEX
    Explanations

    difficult and confusing

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

    The neuron detects words that describe something as difficult, complicated, or challenging (e.g. “hard,” “tricky,” “daunting,” “confusing,” “challenging”).

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    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
    Negative Logits
     kuah
    -1.10
     agus
    -1.02
    :
    -0.96
    oooooooo
    -0.95
     kuma
    -0.93
     kiu
    -0.93
     carlo
    -0.93
     will
    -0.93
    czyk
    -0.92
     usando
    -0.91
    POSITIVE LOGITS
     you
    1.31
    đại
    1.19
    manı
    1.17
    ʏ
    1.11
    Podob
    1.11
    dây
    1.11
    animity
    1.11
    Abraço
    1.09
    Abraços
    1.09
    Especially
    1.09
    Activations Density 0.061%

    No Known Activations