© 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. 74973
    Prev
    Next
    INDEX
    Explanations

    upset victory

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

    This neuron detects mentions of surprising or unexpected outcomes in sports reporting, especially the word “upset” and its variants.

    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
    </h3>
    -2.08
    <bos>
    -1.59
    ”،
    -1.52
    
    -1.41
     Vereinbarung
    -1.33
    geslacht
    -1.28
    You
    -1.25
    اب
    -1.24
    (
    -1.24
    上述
    -1.18
    POSITIVE LOGITS
     in
    1.85
    </b>
    1.84
    ⃙
    1.58
    цію
    1.55
     not
    1.33
     keinginan
    1.31
     verbeteren
    1.30
     sentí
    1.29
    IMENTAL
    1.28
    )\}$
    1.28
    Activations Density 0.018%

    No Known Activations