© 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. 10-GEMMASCOPE-RES-131K
    7. 127348
    Prev
    Next
    INDEX
    Explanations

    expected vs got

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

    The neuron fires on numeric tokens (number literals) wherever they appear.

    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
    </u>
    -1.83
     beiden
    -1.73
     \
    -1.71
    轻易
    -1.62
    ANCHE
    -1.47
     fotográfica
    -1.43
     encuentre
    -1.42
     pestaña
    -1.36
    ۲۱
    -1.35
     كان
    -1.33
    POSITIVE LOGITS
     or
    1.88
     enough
    1.84
     something
    1.71
     not
    1.63
    </h2>
    1.48
     an
    1.47
     more
    1.47
     new
    1.43
     unexpectedly
    1.41
     a
    1.36
    Activations Density 0.013%

    No Known Activations