© 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. 15133
    Prev
    Next
    INDEX
    Explanations

    continue or proceed

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

    This neuron detects polite user‐directed prompts or questions offering choices or actions (e.g. “Would you like to…?”, “Do you want to…?”, “Select this answer”, etc.).

    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
    小时候
    -1.00
     вопро
    -0.99
     apoy
    -0.98
     lorsque
    -0.96
     encuentras
    -0.95
    不时
    -0.94
    柔らか
    -0.93
    看你
    -0.93
     Каждый
    -0.93
     douane
    -0.93
    POSITIVE LOGITS
     it
    1.35
     proceed
    1.15
     or
    1.15
     without
    1.13
     immediately
    1.09
    icidio
    1.09
     continue
    1.03
     PROCEED
    1.01
    блон
    1.01
     further
    0.98
    Activations Density 0.025%

    No Known Activations