© 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. 73465
    Prev
    Next
    INDEX
    Explanations

    programming imports and specific phrases

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

    The neuron fires on capitalized or mixed‐case tokens—that is, proper names, acronyms, class or type identifiers, and other words containing uppercase letters.

    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
     rajut
    -2.52
     hidupnya
    -2.31
    鈇
    -2.23
     actuelles
    -2.19
     officielles
    -2.16
    While
    -2.16
    Even
    -2.14
    '
    -2.13
     barcel
    -2.13
     démocr
    -2.13
    POSITIVE LOGITS
     they
    3.69
     we
    2.64
     largely
    2.47
     his
    2.42
    of
    2.33
     talk
    2.23
     hugely
    2.11
    ⭓
    2.11
     renowned
    2.11
     that
    2.11
    Activations Density 0.000%

    No Known Activations