© 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. 83655
    Prev
    Next
    INDEX
    Explanations

    questions introductions explanations

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

    The neuron strongly responds to tokens containing the vowel “o” (including Cyrillic “о”), with higher activations for tokens having more or repeated occurrences.

    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_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
    觏
    -3.61
    囖
    -3.19
    鼍
    -3.06
    archiwizowane
    -2.98
    
    -2.84
    蒾
    -2.81
    蕺
    -2.78
    actéristi
    -2.78
    毉
    -2.77
    薆
    -2.73
    POSITIVE LOGITS
    ↵↵
    4.25
    м
    3.28
     стал
    2.89
     However
    2.81
     To
    2.80
    $
    2.77
     –
    2.69
    но
    2.67
    …..
    2.64
    \
    2.64
    Activations Density 0.006%

    No Known Activations