© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 98395
    Prev
    Next
    INDEX
    Explanations

    quotes and slogans

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

    The neuron strongly activates on punctuation marks—especially sentence‐boundary or clause‐separating symbols like colons, periods, and em‐dashes.

    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
    cielo
    -1.08
     мечты
    -1.00
     recém
    -0.97
     мебели
    -0.95
     brist
    -0.91
     chamado
    -0.90
     atuação
    -0.90
    ujące
    -0.90
     solche
    -0.89
     его
    -0.89
    POSITIVE LOGITS
     which
    1.09
     motto
    1.05
    这句话
    1.04
     or
    1.03
    rasında
    0.98
     there
    0.97
     می‌گو
    0.96
     lema
    0.95
    ratory
    0.93
     any
    0.90
    Activations Density 0.059%

    No Known Activations