© 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. 34-GEMMASCOPE-RES-131K
    7. 56981
    Prev
    Next
    INDEX
    Explanations

    range or scope

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

    The neuron fires strongly on very common tokens (e.g. high‐frequency words like “a,” “is,” “has”), punctuation, and numbers, rather than on semantically specific content.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    nikami
    -0.83
    stums
    -0.76
    териал
    -0.74
    telah
    -0.74
    tangentMode
    -0.72
    czak
    -0.71
    Lengths
    -0.71
    紡
    -0.71
    叢
    -0.69
    ִי
    -0.68
    POSITIVE LOGITS
     wider
    4.69
     narrow
    4.47
     broad
    4.44
     narrower
    4.34
     wide
    4.25
     broader
    4.22
     widened
    4.13
     broaden
    4.06
     widen
    4.06
     widest
    4.06
    Activations Density 0.133%

    No Known Activations