© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 86088
    Prev
    Next
    INDEX
    Explanations

    fall to defeat

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

    This neuron detects numeric sports outcomes phrased as a team or player “falling to” or “fell to” a particular score or record.

    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
     propelled
    -1.63
    鱻
    -1.60
    說是
    -1.57
    屢
    -1.56
     honn
    -1.56
     soud
    -1.55
     abiye
    -1.53
    鱟
    -1.52
    聃
    -1.52
    废话
    -1.52
    POSITIVE LOGITS
     et
    2.22
     shaky
    1.67
     թվական
    1.65
     ainsi
    1.64
     biasanya
    1.63
     instead
    1.63
     also
    1.62
    $.
    1.60
    parsedMessage
    1.58
     allerdings
    1.57
    Activations Density 0.006%

    No Known Activations