© 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. 39598
    Prev
    Next
    INDEX
    Explanations

    Worldview, projected, outback, View, News, Impressionism

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

    The neuron strongly activates on sequences of uppercase letters or all-caps tokens (e.g. acronyms, headlines, and other fully capitalized strings).

    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
     pomys
    -1.52
     actuales
    -1.48
    från
    -1.46
     mám
    -1.45
    ència
    -1.45
    thschild
    -1.41
     funguje
    -1.41
     compel
    -1.41
     pérdidas
    -1.41
     vélos
    -1.41
    POSITIVE LOGITS
     you
    1.45
    ’
    1.34
    us
    1.23
    tahui
    1.20
     impecable
    1.17
    ⁉
    1.16
    ren
    1.14
     what
    1.13
     recomendable
    1.13
     so
    1.12
    Activations Density 1.590%

    No Known Activations