© 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. 34-GEMMASCOPE-RES-131K
    7. 90515
    Prev
    Next
    INDEX
    Explanations

    bubble bust

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

    The neuron fires on multi‐word prepositional phrases that mark location, time, or numeric context (e.g. “outside the nation’s capital,” “in two batches,” “overseas for the 2020 season”).

    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
     destined
    -0.75
     obstruction
    -0.75
     decomposition
    -0.73
    altet
    -0.73
     larval
    -0.72
    Cooperative
    -0.72
     prestigio
    -0.71
     cooperative
    -0.71
    Transforms
    -0.70
     interception
    -0.70
    POSITIVE LOGITS
     bubble
    2.73
    bubble
    2.38
     bubbles
    2.33
    Bubble
    2.27
     Bubble
    2.20
    bubbles
    1.94
     Bubbles
    1.77
     пузы
    1.52
    Bubbles
    1.47
     BUB
    1.46
    Activations Density 0.005%

    No Known Activations