© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 15403
    Prev
    Next
    INDEX
    Explanations

    traffic stop

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

    The neuron detects tokens referring to a vehicle slowing, stopping, breaking down, or being pulled over.

    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
     但
    -1.56
     也
    -1.54
    Who
    -1.50
     durante
    -1.44
     of
    -1.43
     此
    -1.40
     této
    -1.39
     gesprochen
    -1.38
     داخل
    -1.38
    aimeJ
    -1.37
    POSITIVE LOGITS
     in
    1.86
     légale
    1.54
     nestled
    1.52
     perplexing
    1.50
     provocado
    1.49
    淘宝
    1.47
    urangan
    1.45
     Numerous
    1.45
    Recurring
    1.45
    ساء
    1.45
    Activations Density 0.017%

    No Known Activations