© 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. 47643
    Prev
    Next
    INDEX
    Explanations

    variables like x, y, a, b

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

    The neuron fires on mathematical variable tokens that include numeric subscripts (e.g., x_1, x_2, a_2, b_1).

    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
     نیز
    -0.91
     af
    -0.90
     holdt
    -0.90
     اینترنتی
    -0.89
    nsan
    -0.86
     baada
    -0.85
     moderna
    -0.85
     Sehr
    -0.85
    ましい
    -0.84
    －
    -0.84
    POSITIVE LOGITS
     disponível
    1.04
    뵙
    0.97
     Gründen
    0.93
    總是
    0.92
     conteúdos
    0.88
     órgãos
    0.88
     accél
    0.86
     vlasy
    0.86
     verschillende
    0.85
    ógio
    0.85
    Activations Density 0.012%

    No Known Activations