© 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. 10-GEMMASCOPE-RES-131K
    7. 21147
    Prev
    Next
    INDEX
    Explanations

    toxicological assessment, file identifiers

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

    non-text characters and code syntax elements like fileID, braces, and special Unicode symbols.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron flags rare or garbled non-ASCII symbols (e.g. replacement “�” characters, stray question‐marks, zero-width spaces) in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    opheles
    -2.47
    轶
    -2.45
     ostensibly
    -2.36
    泚
    -2.34
     \
    -2.30
    ↵↵
    -2.28
     is
    -2.23
     unruly
    -2.22
     ajouter
    -2.19
     teclas
    -2.17
    POSITIVE LOGITS
    l
    2.95
    z
    2.70
    y
    2.56
    r
    2.55
    on
    2.53
    ID
    2.48
    p
    2.45
    銪
    2.41
    and
    2.39
    with
    2.39
    Activations Density 0.001%

    No Known Activations