© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 59854
    Prev
    Next
    INDEX
    Explanations

    crash or fall

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

    The neuron strongly activates on terms describing crashes, accidents, or sudden destructive incidents.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     Légion
    -0.86
    ată
    -0.82
    coo
    -0.80
     sputtering
    -0.80
    ストレス
    -0.78
    Решение
    -0.77
    雪花
    -0.77
    itored
    -0.77
    ття
    -0.76
     adventurers
    -0.76
    POSITIVE LOGITS
     crash
    2.69
     crashes
    2.45
     crashed
    2.27
    crash
    2.09
     crashing
    2.09
    Crash
    2.05
     Crash
    2.02
    cras
    1.79
     Cras
    1.57
     falls
    1.38
    Activations Density 0.028%

    No Known Activations