© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 73827
    Prev
    Next
    INDEX
    Explanations

    precedes "of"

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

    This neuron fires on formal, abstract nouns and phrases (e.g. “boundaries,” “accuracy,” “upper limit,” “relationship,” “utilization,” “formation,” “analysis”) that serve as discourse markers in scientific or technical writing.

    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
    la
    -2.70
    i
    -2.47
    n
    -2.33
    of
    -2.28
    —
    -2.27
    In
    -2.20
    )
    -2.20
    us
    -2.20
    They
    -2.19
    to
    -2.13
    POSITIVE LOGITS
    archiwizowane
    1.98
    炁
    1.97
    Ꮠ
    1.96
    zweck
    1.93
     Главная
    1.87
    
    1.87
     and
    1.85
    sandalias
    1.82
    tedì
    1.82
    đầm
    1.81
    Activations Density 0.097%

    No Known Activations