© 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. 100103
    Prev
    Next
    INDEX
    Explanations

    mixed languages and code comments

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

    the digit "0" when it appears as part of a year in the 2000s (like "2005", "2000", "2017").

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5
    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
     —
    -1.62
     after
    -1.61
     because
    -1.42
    $
    -1.42
     this
    -1.41
     {
    -1.41
     if
    -1.38
     while
    -1.34
     mistress
    -1.33
    4
    -1.31
    POSITIVE LOGITS
     étrang
    1.63
     mít
    1.55
     rése
    1.54
     torba
    1.51
     mať
    1.48
    隳
    1.48
     asfal
    1.48
    lemme
    1.47
    鶺
    1.46
     dát
    1.45
    Activations Density 0.000%

    No Known Activations