© 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. 128658
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on Javadoc‐/PHPDoc‐style block comment markers and their leading asterisks (i.e. the “/**” and “*” lines in a doc comment).

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

    javadoc comments /**

    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
     even
    -1.05
     behauptet
    -0.98
     still
    -0.98
     larger
    -0.90
     constructing
    -0.90
     Although
    -0.89
     those
    -0.89
     make
    -0.89
     people
    -0.89
     Even
    -0.88
    POSITIVE LOGITS
    怎么了
    1.09
    𝐌
    1.05
     levert
    1.05
     zoude
    1.03
     manier
    1.03
     verticale
    1.02
     frontale
    1.01
    缑
    1.01
    🔵
    1.01
     mondi
    1.00
    Activations Density 0.006%

    No Known Activations