© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 34580
    Prev
    Next
    INDEX
    Explanations

    foreign language or multilingual context

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

    This neuron reliably activates on running, descriptive prose (the main body text of questions and answers) and stays off for structural markers like “Q:”, blank lines, or purely symbolic/math notation.

    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
     februar
    -0.98
     như
    -0.98
    酵素
    -0.94
     selon
    -0.93
    gü
    -0.90
     januar
    -0.89
    ଂ
    -0.88
     isra
    -0.86
    Mit
    -0.86
     ValidationError
    -0.85
    POSITIVE LOGITS
    only
    0.99
     دیگه
    0.99
     magari
    0.96
     perioada
    0.94
     would
    0.94
    sient
    0.93
    opje
    0.93
    ₂+
    0.92
     irgendwie
    0.91
     Đặc
    0.91
    Activations Density 0.020%

    No Known Activations