© 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. 130414
    Prev
    Next
    INDEX
    Explanations

    This neuron activates on tokens containing the letter sequence “cy” (especially at the start of words, e.g. “Cy…,” “cyto…,” “cyclo…,” “cyano…,” etc.).

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

    Cygwin, Cypher, cyclin, cygnus

    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
    
    -2.80
    更何况
    -2.52
     juzg
    -2.47
    瀦
    -2.47
     are
    -2.44
    brinco
    -2.39
     eigenes
    -2.36
    皞
    -2.33
    这也是
    -2.30
     aswell
    -2.27
    POSITIVE LOGITS
    th
    2.80
     της
    2.45
    All
    2.42
     серь
    2.38
     твои
    2.20
    殲
    2.19
    ',
    2.19
    她的
    2.16
    ｺ
    2.16
    ｵ
    2.14
    Activations Density 0.046%

    No Known Activations