© 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. 10-GEMMASCOPE-RES-131K
    7. 113799
    Prev
    Next
    INDEX
    Explanations

    games, Pane, 方面, GLES

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

    This neuron fires on technical or domain-specific terms and identifiers—words like API names, camel-case or underscored code identifiers, specialized jargon, and proper nouns used in technical contexts.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    '
    -3.84
     you
    -2.78
     there
    -2.73
     This
    -2.59
     When
    -2.56
     \
    -2.44
    ↵
    -2.39
     unparalleled
    -2.36
     three
    -2.36
     has
    -2.36
    POSITIVE LOGITS
    
    3.09
    
    2.52
    苼
    2.50
    䜣
    2.45
    鸶
    2.44
    
    2.44
    镞
    2.44
    賔
    2.44
    ⛦
    2.41
     beri
    2.38
    Activations Density 0.013%

    No Known Activations