© 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. 70838
    Prev
    Next
    INDEX
    Explanations

    two years hours terms

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

    The neuron strongly activates on numeric expressions (both spelled-out numbers like “two” and digit tokens like “3”).

    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
    {
    -2.44
    皤
    -2.20
     serupa
    -1.80
    
    -1.73
     unido
    -1.70
    ㈯
    -1.70
     parachoque
    -1.69
     eletr
    -1.68
    ῞
    -1.68
     surprising
    -1.67
    POSITIVE LOGITS
    ”,
    2.31
    ges
    2.19
    ):
    2.17
     or
    2.11
     but
    2.08
    ;
    
    1.94
     bijvoorbeeld
    1.94
    for
    1.94
    (),
    1.93
    );
    
    1.87
    Activations Density 0.029%

    No Known Activations