© 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. 65765
    Prev
    Next
    INDEX
    Explanations

    separator lines

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

    The neuron activates on long lines composed of repeated punctuation (e.g. hyphens or equals signs) used as section or table separators.

    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
    ꉂ
    -1.46
     Países
    -1.44
     transparan
    -1.37
     ganti
    -1.37
     rusak
    -1.34
    ꦼ
    -1.33
    ırlar
    -1.32
     melap
    -1.28
     dibuka
    -1.28
    gül
    -1.27
    POSITIVE LOGITS
    an
    1.54
    is
    1.47
    al
    1.45
    H
    1.45
    と言う
    1.33
    no
    1.32
    T
    1.27
    i
    1.26
    us
    1.23
     is
    1.23
    Activations Density 0.016%

    No Known Activations