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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 81303
    Prev
    Next
    INDEX
    Explanations

    grace or graves, followed by names/places

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

    The neuron fires on tokens beginning with the substring “Gra” (e.g. Grace, Graves, Gracilibacillus, etc.).

    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
    itivo
    -0.87
     Chippewa
    -0.84
     متوسط
    -0.84
    emplare
    -0.82
    Glorious
    -0.78
    páns
    -0.75
     Citroen
    -0.75
     biv
    -0.74
     टी
    -0.74
     meh
    -0.72
    POSITIVE LOGITS
    茧
    0.86
     Lat
    0.71
     بیم
    0.69
    sug
    0.69
    Lat
    0.69
     ldc
    0.68
    Hs
    0.68
     petition
    0.68
    )!
    0.68
     préf
    0.67
    Activations Density 0.053%

    No Known Activations