© 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. 104387
    Prev
    Next
    INDEX
    Explanations

    enclosed phrases

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

    The neuron fires on idiomatic discourse markers that introduce or emphasize a statement—phrases like “at times. Actually that’s not true,” “getting around … is a nightmare,” or “to say the least.”

    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
     there
    -0.91
     for
    -0.90
     makes
    -0.77
     will
    -0.75
     your
    -0.75
    atuor
    -0.75
     other
    -0.73
     Make
    -0.73
     with
    -0.72
     although
    -0.71
    POSITIVE LOGITS
     understatement
    1.52
     understated
    0.99
    どころ
    0.97
     ENTIRE
    0.95
     MAJOR
    0.94
    整个
    0.93
    instellungen
    0.93
    város
    0.92
     Chính
    0.92
     monstruos
    0.92
    Activations Density 0.036%

    No Known Activations