© 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. 73366
    Prev
    Next
    INDEX
    Explanations

    United Nations, Complex phosphorus

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

    news article formatting and stylistic elements like timestamps, parentheses, and bylines.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron is strongly activated by numeric and scheduling tokens—numbers, dates, times, parentheses around show times or statistics.

    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
     in
    -1.07
    عرض
    -0.88
     destacado
    -0.86
     gets
    -0.84
    jednoc
    -0.84
     at
    -0.84
     on
    -0.77
    jęcia
    -0.75
     prison
    -0.75
     pitcher
    -0.75
    POSITIVE LOGITS
     qualunque
    0.89
     nhiệt
    0.89
     cualquier
    0.85
    MessageOf
    0.85
    凭借
    0.84
     racconta
    0.84
     zamanda
    0.83
    pezifische
    0.83
    istika
    0.83
     responsável
    0.82
    Activations Density 0.002%

    No Known Activations