© 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. 22-GEMMASCOPE-RES-131K
    7. 25315
    Prev
    Next
    INDEX
    Explanations

    car brands and specific entities

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

    The neuron fires on specialist or technical jargon—uncommon, content‐heavy tokens like domain‐specific terms, identifiers, and scientific vocabulary.

    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
     from
    -1.20
     October
    -1.13
     several
    -1.13
     M
    -1.13
     On
    -1.12
     penerima
    -1.11
     September
    -1.09
    These
    -1.04
     You
    -1.01
     November
    -0.99
    POSITIVE LOGITS
     citroen
    1.16
     hond
    1.12
     verra
    1.05
     صفح
    1.05
     broek
    1.05
     vro
    1.04
     Niño
    1.04
     sapatos
    1.04
     pét
    1.03
     latine
    1.02
    Activations Density 0.146%

    No Known Activations