© 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. 62429
    Prev
    Next
    INDEX
    Explanations

    but that's not

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

    The neuron fires on the discourse-marker contraction “that’s” (i.e. the tokens “that” + “’s”).

    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
     bizarre
    -1.12
    similar
    -1.03
    weird
    -1.00
    gled
    -0.99
     vermel
    -0.95
     constitue
    -0.95
    particularly
    -0.94
    Teste
    -0.93
    instead
    -0.93
     botines
    -0.93
    POSITIVE LOGITS
     only
    1.97
     not
    1.54
     because
    1.48
     beside
    1.45
     before
    1.37
     just
    1.25
     for
    1.23
     только
    1.19
     merely
    1.13
     assuming
    1.12
    Activations Density 0.021%

    No Known Activations