© 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. 9348
    Prev
    Next
    INDEX
    Explanations

    examples of non-English scripts

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

    The neuron fires on parts of URLs—particularly domain names (e.g. “.org” and the preceding token).

    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
     for
    -1.41
    ,
    -1.36
    ;
    -1.33
    L
    -1.31
    D
    -1.30
    Only
    -1.23
    M
    -1.23
    -1.23
    It
    -1.22
     it
    -1.21
    POSITIVE LOGITS
     attente
    1.52
     patrie
    1.48
     intelek
    1.44
     vermel
    1.38
     maniere
    1.35
     plafon
    1.30
     péro
    1.28
    锃
    1.28
     kompati
    1.28
     demokrat
    1.25
    Activations Density 0.011%

    No Known Activations