© 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. 64976
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions with exponentiation

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

    The neuron is primarily detecting numeric literal tokens (i.e. multi‐digit numbers) in the math expressions.

    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
    واسطة
    -0.86
     estuviera
    -0.77
    karta
    -0.77
     tuviera
    -0.76
    укра
    -0.76
     worldwide
    -0.75
    ソフトウェア
    -0.74
    asantry
    -0.74
     sikker
    -0.74
     yun
    -0.73
    POSITIVE LOGITS
     uitzicht
    0.90
     */
    0.88
    )}+\
    0.83
    新的
    0.81
     упаков
    0.81
     wireType
    0.78
    ensões
    0.78
    __(/*!
    0.77
     alone
    0.76
     zusätzliche
    0.76
    Activations Density 0.037%

    No Known Activations