© 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. 22-GEMMASCOPE-RES-131K
    7. 104360
    Prev
    Next
    INDEX
    Explanations

    misspelled words and corrections

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

    This neuron responds to angle-bracketed markup tokens (e.g. `<live>`, `<stlsaint>`) and similar tag‐like strings.

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     susc
    -0.87
    lında
    -0.85
    Gem
    -0.84
     implique
    -0.84
     inuti
    -0.81
     fré
    -0.81
     
    -0.81
     jugu
    -0.80
    жных
    -0.79
     нынеш
    -0.79
    POSITIVE LOGITS
    sic
    1.72
     spelled
    1.25
     spelling
    1.14
     missp
    1.09
     misspelled
    0.97
     spelt
    0.91
    spelling
    0.90
     correct
    0.90
     never
    0.90
     [
    0.87
    Activations Density 0.043%

    No Known Activations