© 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. 59787
    Prev
    Next
    INDEX
    Explanations

    hard work and ethic

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

    The neuron fires on language expressing effort, especially occurrences of “hard” and “work” (e.g. “hard work,” “I work hard,” “works as hard as you do”).

    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
    違います
    -0.84
    plevel
    -0.81
    <0x81>
    -0.81
    ډ
    -0.80
     or
    -0.80
     expertly
    -0.77
     Pinus
    -0.77
     Cray
    -0.76
     courage
    -0.74
     ventr
    -0.74
    POSITIVE LOGITS
     ethic
    1.53
     work
    1.46
    work
    1.27
     harder
    1.24
     hard
    1.13
     WORK
    1.09
     Work
    1.08
     travaillent
    1.08
    Work
    1.05
    工作
    0.99
    Activations Density 0.019%

    No Known Activations