© 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. 35791
    Prev
    Next
    INDEX
    Explanations

    more than, more rapidly, more confidently

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

    The neuron is especially responsive to high-frequency English function words (e.g. “more,” “than,” “this,” “one,” “for,” “a”).

    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
     of
    -1.63
    は
    -1.49
     who
    -1.38
    では
    -1.15
    があります
    -1.14
     gioielli
    -1.14
    oxidase
    -1.13
     beve
    -1.11
     consé
    -1.09
    ที่
    -1.08
    POSITIVE LOGITS
    cibly
    1.52
    matically
    1.45
    感じが
    1.40
    与此
    1.37
    lessly
    1.35
    達の
    1.35
    ができます
    1.32
    还得
    1.30
    為に
    1.30
    heartedly
    1.24
    Activations Density 0.109%

    No Known Activations