© 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. 55797
    Prev
    Next
    INDEX
    Explanations

    colors and spectral shifts

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

    The neuron is highly sensitive to color words—especially “blue” (and to a lesser extent “red”)—in the text.

    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
    quido
    -0.83
     hurled
    -0.73
     pły
    -0.73
    organizer
    -0.72
     encompass
    -0.72
    üedad
    -0.72
     đá
    -0.71
    hô
    -0.71
    Stal
    -0.70
    reconnect
    -0.70
    POSITIVE LOGITS
    green
    0.87
    Green
    0.84
     Green
    0.82
     Alem
    0.77
    shift
    0.76
    Red
    0.76
     GREEN
    0.76
     redd
    0.73
     Red
    0.73
    स्या
    0.72
    Activations Density 0.043%

    No Known Activations