© 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. 75156
    Prev
    Next
    INDEX
    Explanations

    special characters and foreign words

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

    The neuron flags structural tokens used in formal documents—such as parenthetical abbreviations (e.g. “e.”), section‐ or list‐heading words (“same,” “last”), and other markers of organized scholarly formatting.

    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
     you
    -1.27
     origen
    -0.94
     fizer
    -0.87
     equilibrio
    -0.86
     ofici
    -0.85
    origine
    -0.84
    ोड़
    -0.84
     doubt
    -0.82
     Compet
    -0.82
    нинград
    -0.81
    POSITIVE LOGITS
     īpa
    1.00
     apparaissent
    0.95
     špec
    0.91
     навіть
    0.89
    复制代码
    0.88
     espaços
    0.86
    werf
    0.86
    specifically
    0.85
    étr
    0.85
    Lovely
    0.84
    Activations Density 0.065%

    No Known Activations