© 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. 34-GEMMASCOPE-RES-131K
    7. 31197
    Prev
    Next
    INDEX
    Explanations

    proper nouns and entities

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

    This neuron responds strongly to punctuation symbols used in code (especially hyphens, dots, and slashes), such as those in CSS custom-property names and file-path 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_34/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
    Punjab
    -0.91
     Punjab
    -0.82
    Pittsburgh
    -0.79
     NRW
    -0.79
     Kleidung
    -0.77
    Cel
    -0.77
    Coronavirus
    -0.76
    
    -0.75
     Amino
    -0.75
     Thiên
    -0.75
    POSITIVE LOGITS
    Disclosure
    0.94
    tail
    0.86
    がありますが
    0.84
     Tail
    0.83
     tail
    0.83
    Задача
    0.81
    ัก
    0.78
    headless
    0.76
     пона
    0.75
     atento
    0.75
    Activations Density 0.002%

    No Known Activations