© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 108648
    Prev
    Next
    INDEX
    Explanations

    decorating and furnishing spaces

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

    This neuron strongly responds to punctuation marks (commas, slashes) and high-frequency function words and contractions (to, and, ’ve, wouldn’t, etc.), i.e. markers of clause or sentence structure.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    食べる
    -0.86
    ỡng
    -0.78
    neys
    -0.78
     fhort
    -0.78
    NEY
    -0.77
     verbunden
    -0.76
     useAppContext
    -0.75
     severed
    -0.75
    tten
    -0.75
     truncated
    -0.74
    POSITIVE LOGITS
     outfit
    2.80
     furnish
    2.39
     furnishing
    2.38
     equipping
    2.06
     equip
    2.03
     furnished
    1.95
     equipped
    1.86
    Furn
    1.84
     Outfit
    1.83
     setup
    1.72
    Activations Density 0.046%

    No Known Activations