© 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. 34-GEMMASCOPE-RES-131K
    7. 48019
    Prev
    Next
    INDEX
    Explanations

    punctuation and symbols

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

    This neuron activates on rare or non-alphanumeric punctuation tokens (e.g. semicolons, underscores, asterisks, tilde, caret, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     ostensibly
    -1.37
    There
    -1.27
     through
    -1.27
     out
    -1.27
    (´
    -1.26
     epitom
    -1.24
     one
    -1.24
     from
    -1.23
     like
    -1.23
     helps
    -1.20
    POSITIVE LOGITS
    possibly
    1.63
    perhaps
    1.53
     tumultuous
    1.46
    probably
    1.45
    actually
    1.45
     joyous
    1.40
     bizarre
    1.39
     luscious
    1.38
    eight
    1.38
    cripts
    1.38
    Activations Density 0.015%

    No Known Activations