© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 61076
    Prev
    Next
    INDEX
    Explanations

    first word of clause

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

    This neuron fires most strongly on words that begin new sentences or lines—that is, sentence-initial tokens.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     you
    -2.86
     we
    -2.09
    2
    -1.84
     however
    -1.84
     therefore
    -1.83
    0
    -1.83
    now
    -1.80
     your
    -1.66
     will
    -1.66
    thylamine
    -1.66
    POSITIVE LOGITS
     Ľ
    1.59
    诸多
    1.52
    voer
    1.49
    Filmographie
    1.49
    --;
    
    1.49
     BorderRadius
    1.48
     fruta
    1.48
     Audiodateien
    1.47
     EconPapers
    1.46
    ')}
    1.45
    Activations Density 0.087%

    No Known Activations