© 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. 56688
    Prev
    Next
    INDEX
    Explanations

    JSON object keys and values

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

    The neuron fires on the punctuation and keywords that mark object–property declarations in JavaScript/JSON (especially the colons and commas separating keys and values).

    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
     even
    -1.18
     usually
    -1.02
     still
    -1.02
     almost
    -1.02
     professionelle
    -0.94
     intensely
    -0.93
     particularly
    -0.93
     satte
    -0.91
     IconButton
    -0.91
     veter
    -0.91
    POSITIVE LOGITS
     zamiast
    1.09
     şun
    1.05
     comprenant
    1.03
     invece
    1.02
    新たに
    0.97
     {
    0.96
     توی
    0.94
    aside
    0.92
     místo
    0.92
     offrant
    0.90
    Activations Density 0.020%

    No Known Activations