© 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. 22-GEMMASCOPE-RES-131K
    7. 95826
    Prev
    Next
    INDEX
    Explanations

    CSS display and position

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

    The neuron specifically picks out CSS style declarations—property names (like “position,” “opacity,” etc.), their colons, and numeric values within style blocks.

    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
     of
    -1.21
     which
    -1.16
     exceeded
    -1.12
     including
    -1.11
     whether
    -1.08
     during
    -1.06
     that
    -1.05
     such
    -1.04
    quelize
    -1.02
     Each
    -1.00
    POSITIVE LOGITS
     normaux
    1.01
    ковая
    1.00
     laaj
    0.98
     habari
    0.98
     colorés
    0.97
     blazer
    0.90
     quilted
    0.89
    I
    0.88
     Système
    0.88
     Ordre
    0.87
    Activations Density 0.016%

    No Known Activations