© 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. 34-GEMMASCOPE-RES-131K
    7. 82929
    Prev
    Next
    INDEX
    Explanations

    attack, swift knee, straight sword

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

    This neuron flags highly technical or data-heavy terms—especially measurement units, quantities, and jargon-style tokens.

    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
     Donau
    -0.78
     invasión
    -0.75
     р
    -0.75
    toolTip
    -0.72
     conoc
    -0.71
    🥍
    -0.69
    ifolium
    -0.69
     мер
    -0.66
    sendEmail
    -0.66
     Uber
    -0.65
    POSITIVE LOGITS
     punch
    6.31
     punches
    5.75
    punch
    5.34
     Punch
    5.19
    Punch
    5.06
     punched
    4.91
     punching
    4.88
     PUNCH
    4.19
    拳
    2.69
     fists
    2.55
    Activations Density 0.071%

    No Known Activations