© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128049
    Prev
    Next
    INDEX
    Explanations

    The neuron detects Java class‐declaration tokens, especially the “extends” and “implements” keywords and the class names that follow them.

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

    java class names

    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
     Dalam
    -0.96
     estran
    -0.94
    鷭
    -0.92
     препарат
    -0.90
     domen
    -0.89
     prostitu
    -0.89
     misterioso
    -0.89
     atras
    -0.85
     MAYOR
    -0.85
    tructive
    -0.85
    POSITIVE LOGITS
     and
    1.02
     with
    0.92
     here
    0.91
     this
    0.87
     but
    0.85
     нашей
    0.81
     этого
    0.79
    libur
    0.77
     apapun
    0.75
     hechas
    0.73
    Activations Density 0.006%

    No Known Activations