© 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. 63511
    Prev
    Next
    INDEX
    Explanations

    given or specified

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

    The neuron is tuned to spot very high‐frequency function words—especially articles and common small “stop” words like “the,” “a,” “for,” “by,” and “when.”

    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
     dezelfde
    -0.90
     rappelle
    -0.89
     iliy
    -0.89
     ấy
    -0.87
    同様
    -0.86
     dasselbe
    -0.86
    鎊
    -0.86
     njia
    -0.85
     utafiti
    -0.85
     costumbre
    -0.85
    POSITIVE LOGITS
     given
    3.66
     specified
    3.02
    given
    2.59
    指定
    2.25
    specified
    2.16
    Given
    2.14
     Given
    1.96
     indicated
    1.90
     provided
    1.89
    Specified
    1.73
    Activations Density 0.091%

    No Known Activations