© 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. 10-GEMMASCOPE-RES-131K
    7. 38377
    Prev
    Next
    INDEX
    Explanations

    woke up or XRD analysis

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

    The neuron fires on rare or specialized tokens—i.e. low-frequency words (such as “woke” or “XRD”) rather than common everyday words.

    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_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
    ?
    -2.69
     When
    -2.48
     zahlreiche
    -2.31
     During
    -2.28
    ).
    -2.23
    十分钟
    -2.14
     pleasant
    -2.13
    ｵ
    -2.09
    ､
    -2.09
     nacer
    -2.05
    POSITIVE LOGITS
    俢
    2.41
    咵
    2.39
    蠵
    2.36
     gän
    2.33
    Polizei
    2.25
     nothwendig
    2.19
    Krankheitsbild
    2.17
     bein
    2.17
    ated
    2.14
     gewiß
    2.14
    Activations Density 0.002%

    No Known Activations