© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 106321
    Prev
    Next
    INDEX
    Explanations

    working and its common collocations

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

    The neuron strongly activates whenever it encounters the word “working” (whether standing alone or as part of compounds like “working-class”).

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    igte
    -0.90
     LETTER
    -0.88
     TICKETS
    -0.86
     prêt
    -0.86
    Lesen
    -0.84
    indd
    -0.84
    FOLLOW
    -0.84
     jobbet
    -0.83
     passende
    -0.83
     senden
    -0.82
    POSITIVE LOGITS
     working
    1.55
     Working
    1.48
    Working
    1.31
    working
    1.22
     knowledge
    1.16
    WORKING
    1.13
     conditions
    1.09
     class
    1.01
     pengetahuan
    0.98
     relationships
    0.98
    Activations Density 0.029%

    No Known Activations