© 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. 10-GEMMASCOPE-RES-131K
    7. 67226
    Prev
    Next
    INDEX
    Explanations

    specific term followed by descriptor

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

    The neuron is detecting uppercase or mixed‐case tokens that serve as names, acronyms, product titles, or other identifiers.

    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
    While
    -1.67
    There
    -1.63
    Since
    -1.59
    Regardless
    -1.59
    After
    -1.56
    Using
    -1.55
    This
    -1.53
     мы
    -1.53
     termurah
    -1.52
    핌
    -1.51
    POSITIVE LOGITS
     they
    2.77
     it
    1.96
     or
    1.75
     was
    1.73
     are
    1.69
     könnten
    1.63
     in
    1.63
     on
    1.55
     even
    1.53
     their
    1.50
    Activations Density 0.272%

    No Known Activations