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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 104205
    Prev
    Next
    INDEX
    Explanations

    pronoun followed by verb or noun

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

    The neuron activates on numeric tokens—especially explicit numbers (including decimals) in the text.

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     it
    -2.09
     in
    -1.84
     and
    -1.84
     on
    -1.70
     for
    -1.64
     These
    -1.63
     That
    -1.38
     whether
    -1.33
     Seems
    -1.31
     Some
    -1.30
    POSITIVE LOGITS
    푉
    1.61
     немає
    1.48
    蓠
    1.45
    xaml
    1.41
     мои
    1.39
    très
    1.38
     migli
    1.38
     нашей
    1.37
    だいぶ
    1.34
     trouvent
    1.33
    Activations Density 0.020%

    No Known Activations