© 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. 33411
    Prev
    Next
    INDEX
    Explanations

    punctuations and reflexive pronouns

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

    The neuron is triggered by formatting‐style tokens—especially digits, dates, commas, and other punctuation used in metadata or boilerplate sections.

    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
    Despite
    -1.26
    Have
    -1.13
    There
    -1.13
     Despite
    -1.13
    there
    -1.09
    Through
    -1.09
    Some
    -1.09
     burgeoning
    -1.07
    Two
    -1.07
    Very
    -1.05
    POSITIVE LOGITS
     his
    1.97
     himself
    1.77
     خودش
    1.27
     trám
    1.17
     després
    1.14
     sooo
    1.13
     soooo
    1.09
     també
    1.09
     praticamente
    1.09
     alſo
    1.07
    Activations Density 0.040%

    No Known Activations