© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14403
    Prev
    Next
    INDEX
    Explanations

    actions and observations

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

    The neuron flags content‐bearing “rich” words—especially vivid action verbs and distinctive descriptive adjectives—amid otherwise functional 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_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
     it
    -0.92
    ストレッチ
    -0.88
     first
    -0.87
     ersten
    -0.86
    instead
    -0.84
    first
    -0.82
    最初に
    -0.81
     cannot
    -0.77
     must
    -0.77
     has
    -0.77
    POSITIVE LOGITS
     then
    1.24
    Якщо
    1.03
    샀
    0.99
    Життє
    0.96
     кстати
    0.96
    Sejarah
    0.95
    StatusEnum
    0.94
    จริงๆ
    0.94
     übrigens
    0.93
     tourner
    0.90
    Activations Density 0.003%

    No Known Activations