© 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. 34-GEMMASCOPE-RES-131K
    7. 117774
    Prev
    Next
    INDEX
    Explanations

    multi-lingual many multiple

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

    The neuron fires on Chinese tokens conveying “many” or “multiple” (e.g. the character 多 and its compounds).

    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
     maksimum
    -0.82
    Nevermind
    -0.80
    🕗
    -0.79
     TAYLOR
    -0.78
     этих
    -0.78
    Thomson
    -0.78
     these
    -0.77
    這些
    -0.76
    ————————————
    -0.76
     dekor
    -0.75
    POSITIVE LOGITS
     sclerosis
    1.16
     years
    1.05
    facet
    0.99
     maraming
    0.98
     prong
    0.94
    times
    0.93
     多
    0.92
     times
    0.91
    Много
    0.91
    choice
    0.90
    Activations Density 0.005%

    No Known Activations