© 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. 22-GEMMASCOPE-RES-131K
    7. 65799
    Prev
    Next
    INDEX
    Explanations

    variable names or code syntax

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

    The neuron flags relatively rare or unusually specific tokens—words and symbols that occur infrequently in the text (e.g. “but,” “not,” “alone,” and digits or mathematical symbols).

    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
    XDD
    -1.20
    reibe
    -1.13
     balení
    -1.09
    UTI
    -1.08
     nových
    -1.01
    歌詞
    -1.00
     Severus
    -1.00
     suppos
    -1.00
     토
    -0.99
     UTI
    -0.99
    POSITIVE LOGITS
     without
    1.10
     in
    1.07
     all
    0.95
     before
    0.93
     four
    0.92
    the
    0.91
     under
    0.90
     using
    0.90
     full
    0.88
     to
    0.87
    Activations Density 0.069%

    No Known Activations