© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 68366
    Prev
    Next
    INDEX
    Explanations

    airport discussions

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

    The neuron mainly activates on the word “airport,” effectively detecting mentions of airports in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    .”
    -1.85
    ‘
    -1.84
    怎么可能
    -1.77
    말
    -1.75
    ’.
    -1.73
     freaking
    -1.72
     admittedly
    -1.72
    addCriterion
    -1.71
    尤其是
    -1.70
    确实是
    -1.68
    POSITIVE LOGITS
    2.14
    戰爭
    2.13
     装修
    2.11
    ㈯
    2.06
    逦
    2.05
    𓆜
    2.02
     répon
    2.02
     itſelf
    1.95
     akal
    1.95
     kosme
    1.94
    Activations Density 0.006%

    No Known Activations