© 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. 10-GEMMASCOPE-RES-131K
    4. 11386
    Prev
    Next
    INDEX
    Explanations

    different languages and symbols

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

    The neuron identifies question boundaries by detecting question‐mark punctuation (i.e. it fires strongly on “?” in question headers).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     a
    -2.61
    ;
    -2.52
     .....
    -2.22
    ........
    -2.20
    .....
    -2.14
    .
    -2.13
     has
    -2.08
     Having
    -2.06
    With
    -2.00
    ’
    -2.00
    POSITIVE LOGITS
     trám
    2.44
     местных
    2.30
    
    2.25
     monumentos
    2.05
     horarios
    2.00
    一个是
    1.98
     เป็น
    1.98
     мужская
    1.97
     nivå
    1.96
     разных
    1.94
    Activations Density 0.063%

    No Known Activations