© 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. 22-GEMMASCOPE-RES-131K
    7. 35700
    Prev
    Next
    INDEX
    Explanations

    locations and dates

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

    The neuron detects the final content words (and punctuation) at the end of a segment or sentence—i.e. tokens appearing at the close of headings, phrases, or sentences.

    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
     kapital
    -0.97
    ichtete
    -0.92
    UNIQUE
    -0.90
     Barbe
    -0.89
    cive
    -0.88
     berge
    -0.87
     Externe
    -0.85
    udz
    -0.85
     PREFACE
    -0.85
    を出
    -0.85
    POSITIVE LOGITS
     clogged
    0.88
    There
    0.80
     already
    0.79
    yd
    0.76
     моих
    0.75
    jąc
    0.75
     at
    0.73
    ẑ
    0.73
    軋
    0.72
     (
    0.71
    Activations Density 0.018%

    No Known Activations