© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129717
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on semantically important content words—mainly nouns and adjectives—rather than on common function words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    both

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    Not in Any Lists

    No Comments

    Negative Logits
    要想
    -0.86
     erstmals
    -0.85
    RatingBar
    -0.85
     específicamente
    -0.84
     các
    -0.82
     проє
    -0.82
     ausdrücklich
    -0.82
    QPushButton
    -0.81
     such
    -0.80
     ル
    -0.80
    POSITIVE LOGITS
     both
    4.22
     både
    3.88
     sowohl
    2.97
    both
    2.95
     zarówno
    2.86
     zowel
    2.73
     BOTH
    2.61
     nejen
    2.59
    BOTH
    2.56
    Both
    2.48
    Activations Density 0.129%

    No Known Activations