© 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. 65245
    Prev
    Next
    INDEX
    Explanations

    relatively followed by adjective

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

    The neuron is tuned to pick out the adverb “relatively.”

    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
    และ
    -2.72
    があり
    -2.20
    aneity
    -2.08
    熜
    -2.06
    
    -2.05
    蠃
    -1.97
    を
    -1.95
    鶿
    -1.94
    そして
    -1.92
    퐹
    -1.92
    POSITIVE LOGITS
    ところに
    2.39
    ’
    2.23
    為に
    1.82
    It
    1.79
    However
    1.72
    場所で
    1.70
     \
    1.69
    While
    1.67
    ままで
    1.66
     llenos
    1.66
    Activations Density 0.017%

    No Known Activations