© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128562
    Prev
    Next
    INDEX
    Explanations

    The neuron is picking out numeric literals (numbers) in the code or JSON.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    filename, scale, or id

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     ideen
    -1.30
     Não
    -1.23
    姽
    -1.20
    liege
    -1.15
     灰色
    -1.14
     kommende
    -1.13
     ferien
    -1.11
    揶
    -1.07
     himmel
    -1.06
    gène
    -1.05
    POSITIVE LOGITS
     may
    1.07
     attaining
    1.03
     or
    1.02
     might
    1.00
     emphasized
    1.00
     customized
    0.99
     emphasizes
    0.98
     had
    0.97
     emphas
    0.97
     sogar
    0.97
    Activations Density 0.006%

    No Known Activations