© 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. 22-GEMMASCOPE-RES-131K
    7. 12759
    Prev
    Next
    INDEX
    Explanations

    concepts and actions

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

    The neuron fires on the central topic word (usually a key noun) that appears at the start of a new section or major paragraph, effectively picking out section‐heading–style content words.

    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
     jejich
    -1.05
     précédentes
    -0.94
    手绘
    -0.93
    Etape
    -0.93
    チール
    -0.91
     éxitos
    -0.91
    Thurs
    -0.91
    eas
    -0.90
     pájaros
    -0.90
    ⼀
    -0.90
    POSITIVE LOGITS
     them
    1.59
     him
    1.37
     his
    1.20
    着
    1.02
    kebox
    1.00
     an
    1.00
    ing
    0.99
     any
    0.97
     to
    0.95
     into
    0.91
    Activations Density 0.145%

    No Known Activations