© 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. 73648
    Prev
    Next
    INDEX
    Explanations

    specific proper nouns and terms

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

    The neuron strongly activates on isolated punctuation or single‐character numeric tokens—i.e. standalone digits and quotation/apostrophe marks.

    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
    atik
    -0.85
    hadur
    -0.84
     serio
    -0.84
    miR
    -0.83
    CardBody
    -0.81
    每个人
    -0.81
     lähe
    -0.81
    唑
    -0.80
     تغذ
    -0.79
     liberality
    -0.79
    POSITIVE LOGITS
     ziyaretçi
    0.94
    ירות
    0.90
    боры
    0.87
     تاريخ
    0.84
    rador
    0.83
     zawsze
    0.83
     precisar
    0.82
    Erst
    0.81
    Ethio
    0.81
    却是
    0.78
    Activations Density 0.005%

    No Known Activations