© 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. 31884
    Prev
    Next
    INDEX
    Explanations

    smiles and laughter

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

    The neuron overwhelmingly fires on words and names having to do with smiling or laughter (e.g. “smile,” “smiles,” “smiled,” “laugh”).

    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
     its
    -0.82
    内で
    -0.78
    ากาศ
    -0.76
    が発
    -0.75
     beginning
    -0.74
     about
    -0.74
     chiefly
    -0.73
    ******/
    -0.73
     solemnly
    -0.70
    adine
    -0.70
    POSITIVE LOGITS
     Laugh
    1.02
     laughter
    0.99
     rire
    0.94
    Laugh
    0.93
    HAHAHAHA
    0.92
    haha
    0.91
     comedy
    0.90
     Laughter
    0.90
    hahahaha
    0.90
     Hil
    0.90
    Activations Density 0.007%

    No Known Activations