© 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. 83013
    Prev
    Next
    INDEX
    Explanations

    point

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

    The neuron is primarily responsive to unusually long or complex tokens (very long words or identifiers).

    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
     betale
    -0.83
    一只
    -0.76
     skrive
    -0.73
    訳
    -0.73
     epo
    -0.72
    Detta
    -0.71
     slutet
    -0.69
    usione
    -0.69
     汁
    -0.68
    ukka
    -0.68
    POSITIVE LOGITS
    point
    2.72
     point
    2.59
     points
    2.36
    points
    2.30
    Point
    2.25
     Point
    2.20
    POINT
    2.19
     POINT
    2.08
     пунк
    1.93
     POINTS
    1.93
    Activations Density 0.036%

    No Known Activations